Grazing Systems to Retain and Redistribute Soil Phosphorus and to Reduce Phosphorus Losses in Runoff
Round 1
Reviewer 1 Report
Below are suggested changes to incorporate in the manuscript along with a few questions. I tried to indicated the corrections with strikeouts and corrections in a red font. I could not see how to upload my Word document, so the red font has disappeared in the copy I have pasted in here. Hopefully, you can figure them out.
Overall, I thought the research question was both interesting and of value. It might be a good exercise to work up a version of the study for non-technical readers who might want to be apprised of its results. I don’t know many farmers who are familiar with analytical spectroscopy and ANOVA statistical tests. But I’m guessing that many would be interested in learning about the two grazing strategies. As a resident on the upper reaches of the Columbia River system, management of phosphorus from non-point agricultural sources is important to the health of our river and its prized trout fishery.
The one item that would help me interpret the results more clearly is an explanation of the AB, Aa, AB etc. label system used in the tables. I was not quite clear as to which letter went with which datum. I did appreciate the figures of the two research center pastures. A lot of information was presented in an understandable and concise form.
Lines 11-17 Much too long an introductory sentence! Break it into at least three parts.
Line 17 comma needed: After three years of treatment, M1P had significantly
Lines 39-41 Reverse this sentence structure: Management practices that…are stocking rate…
Line 48 research have has shown
Line 64 Dung produced by animals 65 in pastures are is deposited
Line 72 They however concluded, however, that
Line 78 with a fenced riparian buffer
Lines 91-95 Switch the order of the two research units to match the order in which their diagrams appear in Figure 1.
Line 96 at two locations, and their respective area coverage 97, and number of
Line 103 Switch order of research center entries in Table 1 to match Figure 1 order.
Line 106 dominant. and Iredell…
Line 120 However, hay Hay, however, was fed…
Line 133 availability. and s Shade and waterers were moved…
Line 134 to the edge of fields or streams.
Line 139 samples points
Line 146 to the lab
Line 150 Mehlich-1 phosphorus (M1P), which is a measure of plant available P, was calculated
Line 151 obtained by using a double acid extractant from a two-component acid mixture.
Line 152 Five grams of each soil sample was were mixed
Line 152 Use subscripts in formula: H2SO4 H2SO4
Line 154 using employing the molybdenum blue dye procedure using with spectrophotometric…
Line 159 with on a Tecan (model number needed here)
Line 169 before the beginning
Line 175 points on within the study pastures
Line 180 space needed: 5-10, and
Line 180 Are the tools in-house software or from a software vendor? If from a vendor, they need to be identified.
Line 183 “Raster Calculator” or A raster calculator?
Line 185 using the Hot Spot Analysis tool
Line 187 hence 55 m
Line 188 Hotspot The Hot Spot tool
Line 196 using by the Wilcoxon Test, which compared median M1P concentrations.
Line 198 analyzed using by the Wilcoxon Test
Line 201 using by the Wilcoxon Test
Line 204 using by the Wilcoxon Test
Line 207 using the JMP software
Line 212 6.1 and 4.9 times increases in median M1P in for the 0-5 cm layer and 2 and 1.6 times increases in median M1P at for the 5-10 cm soil layer
Line 215 inorganic P at in the top 5 cm soil layer for a wide range of soils
Line 216 at in the top 7.5 cm soil layer in of a silt loam soil
Line 218 We speculate, that soil biology
Line 219 to the 5-10 cm interval.
Line 221 Capital needed: m Mehlich-1
Line 222 Table 2 is confusing in its use of the letters following the values. At this first occurrence, they need to be explained in a little more detail. What do the A, a, B, b represent? I couldn’t be sure as a reader. Footnote numbers are missing from table entries as found in Table 3.
Line 225 treatments at on the same sampling date. Differences are at the 0.1 level
Line 229 at the Eatonton pastures
Line 234 eaten in hay and forage back to the soil.
Line 238 With baseline hotspot analysis,
Line 239 at in low-lying areas which that had
Line 247 at the Eatonton pastures
Line 250 at the Watkinsville pastures
Line 252 for the Eatonton
Line 255 pastures, denoted by yellowish to red color in the elevation model, showed greater
Line 261 in the soil in the forms of as DRP and TKP. as t These areas…
Line 263 pastures. , however, STR pastures, due to more uniform distribution of change (no hotspots), presented …
Line 267 five-fold increase in M1P from Baseline to Post-Treatment in the 0-5 cm layer. , M1P in the 5-10 cm layer was…
Line 270 of the STR pastures
Line 274 Capital needed: m Mehlich-1
Line 278 treatments at on the same sampling date. Differences are at the 0.1 level
Line 282 the concentrations, and corresponding loads
Line 284 Comma needed: While these increases are not large, the DRP
Line 290 were noted in Post-Treatment data.
Line 292 evidenced revealed
Line 294 that, even during extreme events, the STR grazing system
Lines 307-309 The exponents for denominator units need to be superscripted.
Line 312 sampling periods. Different letters denote significance at the 0.05 level
Line 315-316 Figure citations seem premature here. The figure is introduced more fully in the next paragraph. I would remove these here.
Line 318 In case of As for DRP loads,
Line 319 a simple regression model. , n No 320 difference in the slopes were…
Line 314 of slope of regression regression slopes
Line 326 Loads of DRP in runoff water was were
Line 328 both treatments at during the Baseline and Post-Treatment samplings.
Line 329 suggest the source of total P
Line 331 This sentence is awkward in construction. Too many “with’s” to sort out.
Lines 334-336 Suggested re-write: More explicitly, areas of high P transport potential need to be managed to reduce P losses in runoff [18]. This includes: 1) steep areas close to streams and prone to erosion, 2) low-lying areas close to the streams, 3) concentrated flow-paths, and 4) high elevation areas with greater slopes.
Line339 into the rhizosphere.
Line 340 down, therefore facilitating
Line 343 vulnerable areas and causing chronic,
Line 344 infiltrate 10 cm into the soil to 10 cm
Line 351 Spaces needed: …treatments. The *, **, and…
Line 355 at the same locations
Line 369 remove hyphen: is needed on farms and beyond the Georgia Piedmont.
Lines 371-372 exponents in denominator units need to be superscripted
Author Response
Dear Reviewer 1:
Thank you for your excellent suggestions. Our responses follow each comment.
Below are suggested changes to incorporate in the manuscript along with a few questions. I tried to indicated the corrections with strikeouts and corrections in a red font. I could not see how to upload my Word document, so the red font has disappeared in the copy I have pasted in here. Hopefully, you can figure them out.
Overall, I thought the research question was both interesting and of value. It might be a good exercise to work up a version of the study for non-technical readers who might want to be apprised of its results. I don’t know many farmers who are familiar with analytical spectroscopy and ANOVA statistical tests. But I’m guessing that many would be interested in learning about the two grazing strategies. As a resident on the upper reaches of the Columbia River system, management of phosphorus from non-point agricultural sources is important to the health of our river and its prized trout fishery.
The one item that would help me interpret the results more clearly is an explanation of the AB, Aa, AB etc. label system used in the tables. I was not quite clear as to which letter went with which datum. I did appreciate the figures of the two research center pastures. A lot of information was presented in an understandable and concise form.
Response: We added examples to improve clarity in each of the tables 2 and 3.
Lines 11-17 Much too long an introductory sentence! Break it into at least three parts.
Response: We broke the sentence into two sentences but were have trouble complying with word number restrictions when we turned into three sentences now lines 11 to 15.
Line 17 comma needed: After three years of treatment, M1P had significantly
Response: Thank you. Comma was added as suggested on line 17.
Lines 39-41 Reverse this sentence structure: Management practices that…are stocking rate…
Response: We reversed the wording structure as suggested. Now lines 40-42.
Line 48 research have has shown
Response: We replaced ‘have’ with ‘has’ as suggested. Now line 49.
Line 64 Dung produced by animals 65 in pastures are is deposited
Response: We replaced ‘are’ with ‘is’ as suggested. Now line 66.
Line 72 They however concluded, however, that
Response: We made changes as suggested. Now line 73.
Line 78 with a fenced riparian buffer
Response: We added ‘a’ as suggested. Now line 79.
Lines 91-95 Switch the order of the two research units to match the order in which their diagrams appear in Figure 1.
Response: The order of the study sites has been switched as suggested on lines 92-96.
Line 96 at two locations, and their respective area coverage 97, and number of
Response: ‘And’ was removed and comma was added as suggested on line 97.
Line 103 Switch order of research center entries in Table 1 to match Figure 1 order.
Response: The order of research center entries has been changed with Eatonton at beginning followed by Watkinsville in Table 1 as suggested to match the Figure 1. Now line 103-105.
Line 106 dominant. and Iredell…
Response: Change has been made as suggested on line 108.
Line 120 However, hay Hay, however, was fed…
Response: The change has been made as suggested on line 123.
Line 133 availability. and s Shade and waterers were moved…
Response: Sentence has been broken down as suggested on line 137.
Line 134 to the edge of fields or streams.
Response: Change has been made as suggested on line 138.
Line 139 samples points
Response: Correction has been made as suggested on line 144.
Line 146 to the lab
Response: ‘The’ has been added as suggested on line 152.
Line 150 Mehlich-1 phosphorus (M1P), which is a measure of plant available P, was calculated
Response: Commas have been added as suggested on line 156.
Line 151 obtained by using a double acid extractant from a two-component acid mixture.
Response: The change was made as suggested on line 157.
Line 152 Five grams of each soil sample was were mixed
Response: The correction has been made on line 158.
Line 152 Use subscripts in formula: H2SO4 H2SO4
Response: The numbers have been changed to sub-scripts as suggested on line 158.
Line 154 using employing the molybdenum blue dye procedure using with spectrophotometric…
Response: Changes were made as suggested on line 160.
Line 159 with on a Tecan (model number needed here)
Response: The change was made as suggested and model number added on line 165.
Line 169 before the beginning
Response: We changed to ‘before runoff begins’ Now line 175.
Line 175 points on within the study pastures
Response: Change has been made as suggested. Now line 181.
Line 180 space needed: 5-10, and
Response: Space has been added. Now line 186.
Line 180 Are the tools in-house software or from a software vendor? If from a vendor, they need to be identified.
Response: We added vendor name “ESRI, Redlands, California” on line 183 at first mention. The vendor name “ESRI” has been added on lines 188 and 193 for the tools used.
Line 183 “Raster Calculator” or A raster calculator?
Response: It is “Raster Calculator” and has been corrected on line 189.
Line 185 using the Hot Spot Analysis tool
Response: “The” has been added. Now on line 193.
Line 187 hence 55 m
Response: “Hence” has been removed from the sentence. Now line 194.
Line 188 Hotspot The Hot Spot tool
Response: Correction has been made as suggested. Now line 196.
Line 196 using by the Wilcoxon Test, which compared median M1P concentrations.
Response: Change made as suggested. Now line 203.
Line 198 analyzed using by the Wilcoxon Test
Response: Change made as suggested. Now line 205.
Line 201 using by the Wilcoxon Test
Response: Change made as suggested. Now line 209.
Line 204 using by the Wilcoxon Test
Response: Change made as suggested. Now line 212.
Line 207 using the JMP software
Response: “The” was added on line 223.
Line 212 6.1 and 4.9 times increases in median M1P in for the 0-5 cm layer and 2 and 1.6 times increases in median M1P at for the 5-10 cm soil layer
Response: Change has been made as suggested. Now on lines 229 and 230.
Line 215 inorganic P at in the top 5 cm soil layer for a wide range of soils
Response: Corrected as suggested. Now on line 232-233.
Line 216 at in the top 7.5 cm soil layer in of a silt loam soil
Response: Changes have been made. Now on line 234.
Line 218 We speculate, that soil biology
Response: Corrected and now on line 235.
Line 219 to the 5-10 cm interval.
Response: “Interval” added as suggested. Now on line 236.
Line 221 Capital needed: m Mehlich-1
Response: ‘M’ was capitalized as suggested. Now on line 238.
Line 222 Table 2 is confusing in its use of the letters following the values. At this first occurrence, they need to be explained in a little more detail. What do the A, a, B, b represent? I couldn’t be sure as a reader. Footnote numbers are missing from table entries as found in Table 3.
Response: We added examples in the footnote of Table 2 in parentheses as (i.e. CHD, 2015 compared to CHD, 2018) to clarify the groups being compared. Now on lines 240-243. Footnotes are changing on upload. We therefore will need help from MDPI and we will reach out to editor upon revision submission.
Line 225 treatments at on the same sampling date. Differences are at the 0.1 level
Response: Change made as suggested. Now on line 242.
Line 229 at the Eatonton pastures
Response: ‘The’ added as suggested. Now on line 246-247..
Line 234 eaten in hay and forage back to the soil.
Response: Change made as suggested on line 252.
Line 238 With baseline hotspot analysis,
Response: Sentence has been corrected. Now line 256.
Line 239 at in low-lying areas which that had
Response: ‘Which’ was replaced with ‘that’. Now line 257.
Line 247 at the Eatonton pastures
Response: ‘The’ added as suggested. Now line 265.
Line 250 at the Watkinsville pastures
Response: ‘The’ added as suggested. Now line 268.
Line 252 for the Eatonton
Response: ‘The’ added as suggested. Now line 270.
Line 255 pastures, denoted by yellowish to red color in the elevation model, showed greater
Response: Commas added as suggested. Now lines 273-274.
Line 261 in the soil in the forms of as DRP and TKP. as t These areas…
Response: Sentences restructured as suggested. Now line 279.
Line 263 pastures. , however, STR pastures, due to more uniform distribution of change (no hotspots), presented …
Response: Sentence has been re-structured as suggested. Now line 281.
Line 267 five-fold increase in M1P from Baseline to Post-Treatment in the 0-5 cm layer. , M1P in the 5-10 cm layer was…
Response: Sentence has been re-structured as suggested. Now lines 286-287.
Line 270 of the STR pastures
Response: ‘The’ added as suggested. Now line 288.
Line 274 Capital needed: m Mehlich-1
Response: ‘M’ was capitalized as suggested. Now on line 292.
Line 278 treatments at on the same sampling date. Differences are at the 0.1 level
Response: Correction has been made in rerunning the statistics we realized it was at the 0.05 level. Now line 294-298. Examples have been added in the footnote to clarify the groups being compared using the letters.
Line 282 the concentrations, and corresponding loads
Response: We made correction as suggested but we also added more results following the same sentence structure as was suggested. Now line 302-303.
Line 284 Comma needed: While these increases are not large, the DRP
Response: With change in post-treatment sample group to individual years for comparison, the sentence now reads: The DRP concentrations and loads for the CHD treatments were significantly greater in 2018 when compared to 2015 and 2017 concentrations and loads (Figure 4). Now line 304-305.
Line 290 were noted in Post-Treatment data.
Response: With changes made in figure 5 because of reanalyzing runoff data, the explanation and values in section 3.5 has changed. Lines 300-326.
Line 292 evidenced revealed
Response: Change has been made as suggested. Now line 312.
Line 294 that, even during extreme events, the STR grazing system
Response: Commas have been added as suggested. Now line 314-315.
Lines 307-309 The exponents for denominator units need to be superscripted.
Response: The exponents of the denominator have been superscripted. Now lines 327-329.
Line 312 sampling periods. Different letters denote significance at the 0.05 level
Response: ‘Period’ has been pluralized as suggested. Now line 332. We have added examples in parentheses in the footnote of the table to further clarify the groups being compared with different letters.
Line 315-316 Figure citations seem premature here. The figure is introduced more fully in the next paragraph. I would remove these here.
Response: The pre-mature figure citations have been removed as suggested. Now lines 335-336.
Line 318 In case of As for DRP loads,
Response: Change has been made as suggested. Now line 338.
Line 319 a simple regression model. , n No 320 difference in the slopes were…
Response: Sentence has been re-structured as suggested. Now lines 340.
Line 314 of slope of regression regression slopes
Response: Change has been made as suggested. Now line 338.
Line 326 Loads of DRP in runoff water was were
Response: Change has been made as suggested. Now line 347.
Line 328 both treatments at during the Baseline and Post-Treatment samplings.
Response: Corrections have been made as suggested by the reviewer. Now lines 349-350.
Line 329 suggest the source of total P
Response: ‘The’ was added as suggested. Now line 350.
Line 331 This sentence is awkward in construction. Too many “with’s” to sort out.
Response: We re-structured the sentence and it now reads, “Transport of different P fractions in runoff and interactions with soil P, in exclusion and non-exclusion areas, provide additional insights on loss and transport mechanisms, especially in areas with high P transport potential.” Now lines 352-354.
Lines 334-336 Suggested re-write: More explicitly, areas of high P transport potential need to be managed to reduce P losses in runoff [18]. This includes: 1) steep areas close to streams and prone to erosion, 2) low-lying areas close to the streams, 3) concentrated flow-paths, and 4) high elevation areas with greater slopes.
Response: Sentence has been re-structured as suggested. Now lines 355-358. We appreciate the excellent suggestion.
Line339 into the rhizosphere.
Response: ‘The’ has been added as suggested. Now line 360.
Line 340 down, therefore facilitating
Response: Comma has been added as suggested. Now line 361.
Line 343 vulnerable areas and causing chronic,
Response: ‘And’ has been replaced with ‘causing’ as suggested. Now line 364.
Line 344 infiltrate 10 cm into the soil to 10 cm
Response: Change has been made as suggested. Now line 366.
Line 351 Spaces needed: …treatments. The *, **, and…
Response: Space has been added as suggested in the footnote of Figure 5. Now line 371.
Line 355 at the same locations
Response: ‘The’ has been added as suggested. Now line 376.
Line 369 remove hyphen: is needed on farms and beyond the Georgia Piedmont.
Response: We rewrote the ending sentence of conclusion and it now reads, “Hence, CHD and STR grazing managements present enormous potential, and further research is needed on farms fertilized with broiler litter and beyond the Georgia Piedmont. Now line 388-390. We added importance of this kind of study in farmers’ fields applied with broiler litter where soil could be high in P and thus pronounced effect on erosion of P in runoff is expected with improved grazing management.
Lines 371-372 exponents in denominator units need to be superscripted
Response: The exponents in the denominator have been super-scripted as suggested. Now lines 392-393.
Reviewer 2 Report
The manuscript quantitatively analyzed the effects of grazing forms (CHD and STR) on phosphorus loss in the watershed in the wild pasture environment, and determined the possible ways of loss. The results of the study are of great significance for understanding the control of grazing methods, reducing the loss of P, and restoring damaged ecosystems. The author had done a lot of work, but in the introduction and discussion, the innovation of the experiment is not well reflected. there are still some areas in the manuscript that need improvement, and the following suggestions are given:
- The introduction was not well written. The conclusion and the purpose of the research is not clear.
- What is the meaning of the example of Line 73-76 "57-83% lower soil macroporosity and 8-17% greater bulk density in areas with free access to cattle compared to areas excluded of cattle were reported" ?
- Add the change of the pasture to support the results obtained in 3.2 and 3.3.
- 3-3.6 only indicated the results without in-depth discussions.
- Which grazing method is adopted in the excluded area in 3.4 ? How to compare with non-excluded area?
- Conclusion section needs to be revised.
Specific comments:
- Line 32-33: How does agricultural use solve the reuse of P in the process of human life ?
- Introduction section: The references are too old. Add the new researches in recent years.
- The picture in section 3.3 does not correct, and there is no corresponding picture on the given website.
- The DRP and TKP methods described in 3.5 are too single, and it is recommended to modify them.
- Line 319: What does slope mean ?
Author Response
Dear Reviewer 2,
We thank you for your comments. We have made changes according to our understanding of your comments. Responses follow each of the comments.
Thank you
The manuscript quantitatively analyzed the effects of grazing forms (CHD and STR) on phosphorus loss in the watershed in the wild pasture environment, and determined the possible ways of loss. The results of the study are of great significance for understanding the control of grazing methods, reducing the loss of P, and restoring damaged ecosystems. The author had done a lot of work, but in the introduction and discussion, the innovation of the experiment is not well reflected. there are still some areas in the manuscript that need improvement, and the following suggestions are given:
- The introduction was not well written. The conclusion and the purpose of the research is not clear.
Response: As per the excellent suggestions made by reviewer 1, the Introduction and the Conclusion are now much clearer.
- What is the meaning of the example of Line 73-76 "57-83% lower soil macroporosity and 8-17% greater bulk density in areas with free access to cattle compared to areas excluded of cattle were reported" ?
Response: We rewrote the sentence to clarify its meaning. It now reads: “Areas with free access to cattle had 57-83% lower soil macroporosity and 8-17% greater bulk density when compared to areas where cattle were excluded [14]”. Now lines 74-76.
- Add the change of the pasture to support the results obtained in 3.2 and 3.3.
Response: We are not sure what Reviewer 2 is requesting but the Change in P maps are in the supplemental materials.
- 3-3.6 only indicated the results without in-depth discussions.
Response: Discussion is in lines 230 to 237 and 244 to 254 for the vertical distributions and 275 to 280 for the horizontal distribution discussions.
- Which grazing method is adopted in the excluded area in 3.4 ? How to compare with non-excluded area?
Response: The strategic grazing. We clarified this in the first two sentences to help the readers understand better. Now lines 285-287. Non-exclusions are all other parts of STR pastures outside of the fenced exclusions.
- Conclusion section needs to be revised.
Response: We have revised the Conclusions with corrections made on line 374-376 and 388-390
Specific comments:
- Line 32-33: How does agricultural use solve the reuse of P in the process of human life?
Response: We have re-written now lines 32-34 with an added reference to clarify the point we are trying to make. The section now reads, “Grazing animals urinate and defecate around 81% of the P eaten [2]. Retention of P in animal waste, within the soils of grazing systems, will improve the efficiency of P use which could contribute to solving this problem.”
- Introduction section: The references are too old. Add the new researches in recent years.
Response: We have added one more reference from 2020 in the introduction on line 32. The old references cited are some seminal works which we think are valuable and wish to retain them. We currently have 15 references that are from 2015-2020. We wish that we had already solved the issues that were identified scores of years ago in those seminal works.
- The picture in section 3.3 does not correct, and there is no corresponding picture on the given website.
Response: The figures referred to in section 3.3 are in the supplemental materials as Figure S1 and Figure S2 which we believe the reviewers should have received along with the manuscript for review. We did submit the supplemental maps in a different file as directed during submission. We need to check with the editor to describe to the reviewers how to access that.
- The DRP and TKP methods described in 3.5 are too single, and it is recommended to modify them.
Response: We have clarified on line 156 that M1P was determined in soil which is different from DRP and TKP determination in runoff water. To make the runoff results clear we redid the analysis of runoff water by sampling years, instead of grouping 2017 and 2018 as Post-Treatment period. Hence, we have made appropriate changes in section 3.5 on lines 300-304, and lines 316-325. Subsequent changes were made in methods on lines 150-152 and lines 210-214, and results in 308-315.
- Line 319: What does slope mean?
Response: We have restructured the sentence to clarify that the slopes mean the regression slopes. The sentence now reads, “As for DRP loads, the regression slopes derived from the relationship, DRP versus M1P (0-5 cm soil depth), during Baseline and Post-Treatment were compared using a simple regression model. No differences in the slopes were indicated for CHD pastures between Baseline and Post-treatment, while the slope in STR pasture was significantly lower for Post-Treatment as compared to Baseline.” Now on lines 338-342.
Reviewer 3 Report
Review of the article entitled "Grazing Systems to Retain and Redistribute Soil Phosphorus and to Reduce Phosphorus Losses in Runoff"
The authors evaluated the effect of two different grazing systems that affect the P losses out of the pasture system. To evaluate the effect of grazing on P loss, two systems were established, 1) strategic-grazing (STR), and 2) continuous-grazing-with-hay-distribution (CHD). The two systems were implemented in two locations, bot with different soils, although mainly alfisols, ultisols and entisols. The authors also describe the vegetation composition of the locations. The experimental design is correctly applied, the soil sampling (grid of 50 m collected at 0-5, 5-10, 10-20 cm depth) and the number of sampling points is proper. Exclusion areas were established in the STR system. The chemical procedures as the statistical ones are described and supported by the literature.
The main results show that the two strategies of grazing reduced the amount of P loss from the system and that P accumulated in the surface soil layer (0-5 cm), this is in relation to the baseline that the authors established in the year 2015, and the measurements that were done in 2018 (Table 2). With the soil and water samples taken in a grid in both locations, the authors built up a series of maps to indicate places susceptible to be more affected by P losses due to the grazing behavior of the cattle (Figure 2 and 3). With these maps, the authors were able to identify sites of accumulations, in principle in places with high elevations, i.e. the P is not lost to depressional places by surface runoff. Overall, the STR retain more P in the soil and prevent more efficiently P losses.
Question: are the slopes in figure 5 correctly? i.e. the points seem to have a large dispersion, and then probably slopes are not significantly different.
Best regards
Author Response
Dear Reviewer 3:
Thank you for your comments. We tried to better describe our statistical methods and have added reference to support them.
Respectfully,
Question: are the slopes in figure 5 correctly? i.e. the points seem to have a large dispersion, and then probably slopes are not significantly different.
Response: We re-ran the statistics and found no difference in results. We also further clarified the statistical method we used to compare the regression lines during Baseline and Post-Treatment in section 2.7 on lines 214-222.
Reviewer 4 Report
The paper “Grazing Systems to Retain and Redistribute Soil Phosphorus and to Reduce Phosphorus Losses in Runoff “ by Anish Subedi et al. evaluates the potential of two grazing systems for reducing phosphorus losses in runoff via the retention and redistribution of phosphorus into the soil. They use different grazing practices to direct livestock grazing away from vulnerable areas and to prevent livestock congregation in those areas. They measured soil P content, and different P fractions in runoff water before and after 3 years of different grazing systems. Results showed that both grazing systems improved the P content in the first 5 cm of soil and that P hot-spots moved from vulnerable localizations to less vulnerable areas. They present interesting results with implications for sustainable management of grasslands.
In general, the manuscript is well written, structured and easy to read and authors deliver a concise and clear message. However, there are a number of issues that need to be clarified/addressed:
- Although the authors provide sufficient background and they correctly justified their study, in my opinion, there is room for improving the introduction. I think that the authors present different ideas in the same paragraph and the introduction would improve stating one principal idea per paragraph. My suggestion is a first paragraph indicating the importance of improving the efficiency of P use within agricultural systems, a second paragraph describing (briefly) the P cycle in grazing systems, a third paragraph commenting how the P use efficiency can be improved in relation with the different processes of the P cycle and a last paragraph stating the objectives of the study.
- Material and Methods. My main concern with this paper is that the authors do not provide enough methodological details for allowing the reader to correctly asses their results. The authors should justify that they can soundly compare pre- and post- grazing treatments. Can the authors attribute the differences between pre- and post-treatment only to the grazing systems? Can other factors have an effect in the observed differences?
I miss more details about the grazing systems (pre and post). I would appreciate a brief description of the intensity of the different grazing systems. For example, are the stocking rates similar (comparing the different systems pre-, CHD y STR and the two different pastures)?
I also miss sampling dates for the soil monitoring. I assume that you only sampled soils once in 2015 and another time in 2018, but when, exactly? Are the sampling dates similar? Are the observed pre- and post- differences due to the grazing systems or a seasonal effect? Could the different inter annual environmental conditions (precipitation, temperature) affect the observed differences? For this reason, I wold appreciate some information about the environmental conditions during the period of study. I think this is especially important when comparing P loads. Differences between baselines and post-treatments are due to the treatments or to differences in precipitation?
- Statistical analysis. Viewing the boxplots of Figure 4, I would assume that the distributions of the variables are skewed. Although the use of Wilcoxon test in this case is correct, I suggest that the authors look at generalized additive models for location scale and shape (GAMLSS), if not for this paper, for future research. GAMLSS allow to model up to four parameters in a distribution family. You could choose a family distribution for variables highly skewed that can model not only the median and a parameter of variation, but also a parameter that measures the skewness and a parameter that measures the kurtosis. These models are more flexible and you could detect differences (or not) not only between medians but also in the skewness of the distributions. If you are an R user you could check the GAMLSS package.
- Results and Discussion. When studying relationships between the M1P in soil and TKP in runoff water (Figure 5) I am not sure how the authors relate a measure of soil P content (samples from the 50X grid and supposedly taken one time) with a measure of a runoff sample (several samples taken each time of a precipitation event). I would appreciate that the authors clarify how they made this relationship. Viewing the figure and the r values of the regressions, I am not so sure that simple linear regression is the best approach. You are comparing slopes of linear functions when sometimes the relationships between the variables appear to be non-linear (at least for CHD post-treatment). Maybe you could check non-linear models.
- L226-L236. I think the authors should revisited and re-think this paragraph. It is clear that CHD received more P in form of hay than STR, so the authors attributed the improvement of soil P in the CDH pastures mostly to the hay addition and in the STR pastures mostly to the retention of P deposited by grazing cattle. But as I see it, the major part of the hay was consumed by the cattle, so the major part of the P added in form of hay ends into the soil after being consumed first by the cattle.
Author Response
Dear Reviewer 4:
We thank you greatly for your comments. They have added more clarity to our manuscript. Our responses follow each of your individual comments.
Thank you,
The paper “Grazing Systems to Retain and Redistribute Soil Phosphorus and to Reduce Phosphorus Losses in Runoff “ by Anish Subedi et al. evaluates the potential of two grazing systems for reducing phosphorus losses in runoff via the retention and redistribution of phosphorus into the soil. They use different grazing practices to direct livestock grazing away from vulnerable areas and to prevent livestock congregation in those areas. They measured soil P content, and different P fractions in runoff water before and after 3 years of different grazing systems. Results showed that both grazing systems improved the P content in the first 5 cm of soil and that P hot-spots moved from vulnerable localizations to less vulnerable areas. They present interesting results with implications for sustainable management of grasslands.
In general, the manuscript is well written, structured and easy to read and authors deliver a concise and clear message. However, there are a number of issues that need to be clarified/addressed:
- Although the authors provide sufficient background and they correctly justified their study, in my opinion, there is room for improving the introduction. I think that the authors present different ideas in the same paragraph and the introduction would improve stating one principal idea per paragraph. My suggestion is a first paragraph indicating the importance of improving the efficiency of P use within agricultural systems, a second paragraph describing (briefly) the P cycle in grazing systems, a third paragraph commenting how the P use efficiency can be improved in relation with the different processes of the P cycle and a last paragraph stating the objectives of the study.
Response: We have re-written the initial paragraph of the introduction and added one more reference to clarify the section. Now lines 32-34. We have also re-structured sentences on lines 40-42 and lines 74-76 to improve the readability of the introduction section.
- Material and Methods. My main concern with this paper is that the authors do not provide enough methodological details for allowing the reader to correctly asses their results. The authors should justify that they can soundly compare pre- and post- grazing treatments. Can the authors attribute the differences between pre- and post-treatment only to the grazing systems? Can other factors have an effect in the observed differences?
Response: We added cattle densities for the study pastures, during Baseline on lines 118-119, and during Post-Treatment on lines 131-132. Cattle densities did change from Baseline to Post-Treatment but did so equally for both treatments. We clarified the soil sampling dates on lines 143-144 of the Methods. The sampling dates were uniform for both treatments. No statistical differences were observed between the two locations, Eatonton and Watkinsville, and all pastures experienced similar environmental conditions during the study periods (Baseline vs Post-Treatment). While discussing the changes in soil P distribution, we have indicated that more hay was fed in the CHD pastures. But please note that two reps of each treatment were located at both sites. The treatments were exposed to the same environmental condition. Thank you for pointing these out. We need to include this information in the manuscript.
I miss more details about the grazing systems (pre and post). I would appreciate a brief description of the intensity of the different grazing systems. For example, are the stocking rates similar (comparing the different systems pre-, CHD y STR and the two different pastures)?
Response: We added cattle densities for the study pastures, during Baseline on lines 118-119, and during Post-Treatment on lines 131-132.
I also miss sampling dates for the soil monitoring. I assume that you only sampled soils once in 2015 and another time in 2018, but when, exactly? Are the sampling dates similar? Are the observed pre- and post- differences due to the grazing systems or a seasonal effect? Could the different inter annual environmental conditions (precipitation, temperature) affect the observed differences? For this reason, I would appreciate some information about the environmental conditions during the period of study. I think this is especially important when comparing P loads. Differences between baselines and post-treatments are due to the treatments or to differences in precipitation?
Response: We added the soil sampling dates on lines 143-144 of the methods. The sampling dates were uniform for both treatments. No statistical differences were observed between the two locations, Eatonton, and Watkinsville.
- Statistical analysis. Viewing the boxplots of Figure 4, I would assume that the distributions of the variables are skewed. Although the use of Wilcoxon test in this case is correct, I suggest that the authors look at generalized additive models for location scale and shape (GAMLSS), if not for this paper, for future research. GAMLSS allow to model up to four parameters in a distribution family. You could choose a family distribution for variables highly skewed that can model not only the median and a parameter of variation, but also a parameter that measures the skewness and a parameter that measures the kurtosis. These models are more flexible and you could detect differences (or not) not only between medians but also in the skewness of the distributions. If you are an R user you could check the GAMLSS package.
Response: Thank you for the excellent input. We will explore this statistical method suggested by the reviewer to explore other possible manuscripts as suggested by the reviewer.
- Results and Discussion. When studying relationships between the M1P in soil and TKP in runoff water (Figure 5) I am not sure how the authors relate a measure of soil P content (samples from the 50X grid and supposedly taken one time) with a measure of a runoff sample (several samples taken each time of a precipitation event). I would appreciate that the authors clarify how they made this relationship. Viewing the figure and the r values of the regressions, I am not so sure that simple linear regression is the best approach. You are comparing slopes of linear functions when sometimes the relationships between the variables appear to be non-linear (at least for CHD post-treatment). Maybe you could check non-linear models.
Response: We have added more detail on the statistical method used to compare the relationships between soil P and runoff P loads on lines 214-224. We plotted the event loads of runoff P (DRP loads, and TKP loads) of each watershed for the average annual soil-P value of the individual watersheds within each of the pastures. We have added a reference to support our analysis methods on line 218.
- L226-L236. I think the authors should revisited and re-think this paragraph. It is clear that CHD received more P in form of hay than STR, so the authors attributed the improvement of soil P in the CDH pastures mostly to the hay addition and in the STR pastures mostly to the retention of P deposited by grazing cattle. But as I see it, the major part of the hay was consumed by the cattle, so the major part of the P added in form of hay ends into the soil after being consumed first by the cattle.
Response: We have added a reference on line 32 of the introduction which highlights the P use efficiency of livestock including grazing cattle. Although the added P in haybales could have been harvested in the form of animal growth, the largest amount of P (~ 81%) ends up in urine and manure. Added hay bales in CHD pastures are adding more P to the CHD pasture system whereas, in STR pastures, the recycling of P appears to be more efficient with less need of external input of P to the system.
Round 2
Reviewer 2 Report
Different grazing systems affect soil nutrients and runoff processes by changing the underlying surface conditions. The migration and loss characteristics of phosphorus in the soil along with surface runoff have important practical and theoretical significance for preventing non-point source pollution and mitigating soil erosion. This study lasted for 3 years. A large number of sample points were arranged in two pastures. The temporal and spatial changes of phosphorus in soil and runoff were compared under two grazing methods (STR and CHD). The research reflects a lot of meticulous work, which is obviously very instructive to actual production practice, but there are some problems in the narrative structure. Give these following suggestions:
Main:
1) In the introduction part, the logic is rather confusing. It is recommended to introduce in sections from several key core issues, such as the effect of grazing on the vertical distribution of soil phosphorus content and research progress. The effect of grazing on soil surface phosphorus content and phosphorus input and output. Research and progress on phosphorus loss from pasture runoff by different grazing systems,etc. Finally, it is necessary to explain the differences or innovations between this study and previous studies.
2) Line 255, it is recommended to change the title. Cluster analysis is just a method of data analysis, and the title should reflect the subject of your work in this part. Perhaps, you can mergr section 3.2 and 3.3, both of which are about the spatial distribution of phosphorus variation characteristics.
3) Sections 3.1 and 3.5 can be mergred. These two sections are actually the characteristics of the vertical distribution of phosphorus. The exclusion area is just a special underlying surface, which can be used as a reference sample series to other areas.
4) Table 2 shows the overall mean values of soil layers at different depths under the two methods (STR and CHD). What is the difference between the vertical distribution of phosphorus in low-lying areas and high-lying areas? Suggest additional related discussions.
Specific:
1) line 20, it should be 21%.
2) The text in Figure 1 is mixed with the sampling point and it is not clear. Please change the color.
3) The format of Table 1 is messy and it is recommended to adjust.
4) Line 106-114, soil information, slope information, etc. can be sorted into tables, which is more concise.
5) In Figures S1 and S2, the red and blue boundary lines appear unclear under high-saturation color matching. It is recommended to bold the border.
6) The ratio of the DEM image in Figure S1, S2 should be consistent with the rest of the images. At the same time, the unit "m" should be added after the DEM legend value.
Author Response
Dear Reviewer 2,
We have responded to your comments and believe they have improved the manuscript. There was one section (3.5) that we did not move. While we did greatly rearrange the Introduction as requested, we believe section 3.5 to be logically placed. In addition the three other reviewers did not have an issue with the location of section 3.5.
Below please find our responses to your comments. You can find our responses below each of your requests and in the revised manuscript that is in track change mode.
Again, thank you for your helpful review.
Respectfully submitted,
Dory Franklin and Anish Subedi
Different grazing systems affect soil nutrients and runoff processes by changing the underlying surface conditions. The migration and loss characteristics of phosphorus in the soil along with surface runoff have important practical and theoretical significance for preventing non-point source pollution and mitigating soil erosion. This study lasted for 3 years. A large number of sample points were arranged in two pastures. The temporal and spatial changes of phosphorus in soil and runoff were compared under two grazing methods (STR and CHD). The research reflects a lot of meticulous work, which is obviously very instructive to actual production practice, but there are some problems in the narrative structure. Give these following suggestions:
Main:
1) In the introduction part, the logic is rather confusing. It is recommended to introduce in sections from several key core issues, such as the effect of grazing on the vertical distribution of soil phosphorus content and research progress. The effect of grazing on soil surface phosphorus content and phosphorus input and output. Research and progress on phosphorus loss from pasture runoff by different grazing systems,etc. Finally, it is necessary to explain the differences or innovations between this study and previous studies.
Response: We have rearranged the Introduction to be more in line with the Reviewer 2 requests. We also added additional references related to vertical and horizontal distribution (mobility) of P. They are now citation numbers [6], and [20].
2) Line 255, it is recommended to change the title. Cluster analysis is just a method of data analysis, and the title should reflect the subject of your work in this part. Perhaps, you can mergr section 3.2 and 3.3, both of which are about the spatial distribution of phosphorus variation characteristics.
Response: We have merged sections 3.2 and 3.4 as requested.
3) Sections 3.1 and 3.5 can be mergred. These two sections are actually the characteristics of the vertical distribution of phosphorus. The exclusion area is just a special underlying surface, which can be used as a reference sample series to other areas.
Response: The exclusions also acted as traps or filters to reduce downhill mobility of P in runoff waters. Placement of section 3.5 just before the runoff water is a logical place for this section. Evidence of the vertical movement of P further into the soil supports our runoff results and ties the two sections together. We therefore would like to keep this section where it is.
4) Table 2 shows the overall mean values of soil layers at different depths under the two methods (STR and CHD). What is the difference between the vertical distribution of phosphorus in low-lying areas and high-lying areas? Suggest additional related discussions.
Response: Table 2 shows median values. Figure 2 elucidates the change in horizontal distribution of P from 2015 to 2018 at different depths. Concentrated flow paths illustrated in Figure 2 are markers which describe where the water is concentrating in the pastures and moving downhill. We did not do analyze differences between high-lying and low-lying areas and believe this manuscript is quite extensive as is. It would be an interesting analysis though beyond the scope of this manuscript.
Specific:
- line 20, it should be 21%.
Response: We added the percent sign. Thank you.
- The text in Figure 1 is mixed with the sampling point and it is not clear. Please change the color.
Response: We changed the color of the treatment and pasture names as requested
- The format of Table 1 is messy and it is recommended to adjust.
Response: We have adjusted Table 1 by better aligning header and columns and better delineating the headers.
- Line 106-114, soil information, slope information, etc. can be sorted into tables, which is more concise.
Response: We added a table as suggested.
- In Figures S1 and S2, the red and blue boundary lines appear unclear under high-saturation color matching. It is recommended to bold the border.
Response: We bolded the borders as suggested. It looks much better now. Thank you.
- The ratio of the DEM image in Figure S1, S2 should be consistent with the rest of the images. At the same time, the unit "m" should be added after the DEM legend value.
Response: We amended S1 and S2 as suggested.