Production Indicators in Cowpea as a Function of the Water–Fertiliser Nexus

Round 1

Reviewer 1 Report

agronomy-575190

“Production indicators in the cowpea as a function of the water-energy-fertiliser nexus”

General summary:

This manuscript investigates the benefits of waste-water use on cowpea production in the northeast of Brazil, comparing four measures of productivity, under two irrigation types (conventional tap-water and alternative treated waste-water) and three contrasting fertilizer regimes (none, organic and mineral). This manuscript is reasonably well written and easy to follow, highlighting the potential benefits of using treated waste-water for irrigation of cowpea, resulting in maintained yields, with no need for additional fertilisers, and producing similar production measures compared with conventional practices of tap-water irrigation and mineral fertiliser applications. There are, however, several aspects of the manuscript that need to be addressed. In particular some conclusions are not clearly supported by the results and should be changed or made more specific. The emphasis on ‘energy’ source for the irrigation system does not appear to clearly relate to the results of this study and should probably be removed from the manuscript and its title. In addition, the manuscript should highlight better, the novel aspects of this study and how it advances the field, compared with previous similar research. The Figures/Tables also need thorough revision, as the same data is presented across several figures, and some are not necessary. I hope the following comments help in revising the manuscript.

Broad comments:

The conclusions need to be more specific. Many of the conclusions are made with no comparison, i.e. L26-27, 214-219, 409-410, “better performance” than what, “increases” compared to what? What is this compared to? Similar conclusions should be revised, specifying in each case, what it is “better” than, or what it “increases” compared with. Stating something is “better” or “increases” does not mean anything if it is not compared to something else. Furthermore, some conclusions do not specify what measurements are being compared, or what results they are based on, i.e. L220-222. Further to the previous comment, please be careful and make sure that all major conclusions are based on results where there is a significant difference in the data. Throughout the ‘Results and Discussion’ there are several major statements that are not supported by the data, or are only partially supported. For example, L230-232 (S1A1 was not different from the S2 treatments), L236-239 (no difference between S2A0 and S2A1 for any measure of productivity), L240-243 (S2A2 only greater in productivity than S1A2/S1A0), L280-281 (A2 not different from A1 for NPP under both irrigation types), L390-392 (not true, no difference), L399-403 (not true, no difference). Most importantly, the conclusion (L409-410, L214-216) that the wastewater treated plants all performed better (presumably compared to the conventional watering treatments) needs to be changed, as they did not all perform better, i.e. the productivity indicators for S2 treated plants were not higher (P>0.05) than under S1A1 (productivity) or S1A1/S1A2 (W100, NGP, NPP) treatments. The statement can instead be made that all wastewater treatments (S2A0, S2A1, S2A2) showed similar yield compared with conventional watering with mineral fertiliser (S1A1), therefore demonstrating that the use of treated waste-water can substantially reduce, if not, replace fertiliser requirements in cowbean production in northeast Brazil. Please revise the above results and conclusions and ensure that all reported results are properly and clearly supported by the data and the statistical analyses (a slightly higher average does not mean one treatment is higher than another, there should be a significant difference in the post-hoc analysis, at P<0.05, otherwise this should be acknowledged with caveats). There appears to be quite a lot of emphasis on the ‘energy’ aspect of the irrigation system; solar versus conventional grid power. For example, it is part of the title “water-energy-fertiliser nexus”, it is presented in detail in the methods (L105-113) and is mentioned as part of the treatment factor regularly throughout the ‘Results and Discussion’, i.e. L214-215: “with wastewater and driven by photovoltaic solar energy”, L328: “For the factors of energy + water and the interaction (energy, water and fertiliser)”. This makes the “energy” source of the irrigation system seem like a major influencing factor in the study. Yet, I do not see how the energy source plays any role in the results of the study, or why it is considered as such a major part of the irrigation factor? I understand it is important to compare “conventional” versus “alternative” irrigation strategies, but to include the term “energy” as a seemingly major part of this is quite misleading and I felt myself expecting this to be addressed in the discussion, perhaps in terms of the cost of production or something, but it never comes up. I would suggest revising this aspect and if it is not going to form a significant part of the study then the emphasis on “energy” source should be removed from the paper. For example, it should be removed from the title, and should only really be discussed in the methods, and not as part of the factor in the ‘Results and Discussion’. I find this research promising for encouraging the use of treated wastewater in agricultural systems and consequently reducing current demand on fertilisers. However, as highlighted by the authors there are many similar studies showing the benefits of using wastewater to increase or substitute for fertiliser inputs, yet maintain productivity/yield. Therefore, the authors need to be clearer in what the novel aspects of this research are? Unfortunately, I feel the Introduction and Discussion do not adequately highlight/emphasise the novel aspects of this research. What exactly are the important novel aspects of this research and how does it represent a significant advancement to this field of science, beyond other similar studies? Some speculative interpretations are presented as fact, and these should be toned down. i.e. L217-219: it is not known if the nutritional requirements of the crop were met, as no plant nutrient analyses were performed. The results certainly suggest this is the case, but the language should be toned down to more accurately reflect the results. i.e. add “likely met the nutritional demand”. Also, this may result in higher nutrient uptake, but not necessarily a greater efficiency in nutrient absorption. This should also be termed differently and again change the wording, as this was not tested in this study and so this is not known for certain. L259: why is this impossible? Please explain further as to why this would be the case? Also, L260-263, L380-381, L392-394, change wording. The same data is presented in multiple figures/tables. For example, Table 7, Figs 3, 4, 5, 6 and 7 all repeatedly show parts of the same data. I would suggest including Table 7 as supplementary material, combining Figs 3-6 in one or two figures and removing Fig 7 (please see more specific comments below). I would suggest a closer analysis of the changes in soil nutrients under the waste-water treatment and a proper discussion of this in the manuscript. In particular phosphorus appears to substantially increase in the soil under waste-water compared with under conventional tap-water; are the availabilities of other nutrients also significantly increased in the waste-water treated soils? This is barely raised at all in the manuscript and is not statistically tested. Similarly, there should be some statistical analysis of soil data, to test for other differences. Is P often limiting to productivity in this system? The term “production components” is used throughout. I think a more appropriate term is the term used in the title, “production indicators”, or alternatively “production measures”. I would suggest changing this throughout the manuscript, including in the Table and Figure captions.

Specific comments:

Abstract:

L17: It would be good to specify that cowpeas are a bean, so that this sentence leads on from the first line, i.e. “…in the bean crop cowpea, under…”

L20: Please remove the acronym “CRD” as is it is not used again.

L21-24: These two sentences are repetitive, it would be better to combine them into a single sentence.

L27: Change to “…for all measures of cowpea production.”

Introduction:

It would be helpful to include a very brief summary of the experimental design and the main measurements in the Introduction. This helps the reader to know what to anticipate and how exactly you will test your hypotheses/aims.

L58: add reference to the second sentence.

L60-63: this is not an increase of 935.8 kg ha^-1, please rephrase this sentence.

Materials and Methods:

L81: change to “…this study was to…”

L86: add “…to the control (no fertiliser).”

L87: change to “…in length, with 42 plants…”

L183-185: how many samples were taken and how were they analysed. Please provide a brief summary of the methods for all soil analyses.

Results and Discussion:

L229: should this be “…under wastewater…”?

L230-232: Acknowledge here that S1A1 was equivalent in productivity to the wastewater treatments.

L240-243: This needs to be supported, either from a reference or with specific results from this study.

L260-263: Wouldn’t this mean that the mineral fertiliser would do the same thing? Considering there is very little to no difference in productivity among all S2 and S1A1 treatments, then this would suggest to me that under these treatments they have all reached their maximum productivity in this system and that there is no additional benefit from extra fertiliser, regardless of whether it is mineral or organic.

L273: but why?

L306-307: I would suggest removing this last sentence, as there is a significant interaction and it can be mis-leading to report just the ‘fertiliser’ difference, when there is a significant interaction.

L325: I assume this is a cowpea cultivar? Please specify.

L334-336: To help make the point stronger here, it might be better to make it clearer how this relates to your study. Similarly, where other studies are used to support the findings I would suggest making it clearer exactly how they relate.

L341: change to “…had the highest mean…”

L358-356: as there was a significant interaction factor in productivity it can be mis-leading to report the overall mean across all fertiliser treatments, instead I would suggest reporting the range of the three fertiliser treatments under the tap-water and waste-water treatments.

Figures and Tables:

Table 2: What does the “Additional” column mean?

Table 3 and 4: I would suggest combining Tables 3 and 4 to allow for a direct comparison of the water properties. What is the “Reference for reuse” column? Please specify if these are mean values and what the n is. Please include the standard errors. Please specify all abbreviations in the Table caption (i.e. NS, EC, COD, SST), also, why are there missing values in the table? For total coliforms I would suggest changing the units to avoid having to use these more extreme values. In Table 4 please change the values for ‘total coliforms’ and ‘Escherichia coli’ to not be 0.0.

Table 5: again, please define all abbreviations in the table caption (other than elements). What is ‘assimilable P’?  State somewhere the unit of depth indicated in the table (i.e. depth (cm) or 0 – 20 cm ect.). Please specify if these are the mean values, from how many replicates and what the standard error is. It would be great to know if there is actually a statistical difference between the ‘start’ and ‘end’, otherwise these values aren’t as useful, as there may or may not be an actual difference there. It is impossible to know without any standard errors or statistical comparison!

Table 6: same comments here, are these means, please add n, se, specify all abbreviations.

Table 7: same comments here, are these means, please add se, specify all abbreviations. I would suggest including this kind of descriptive statistics in the supplementary materials.

Table 8-10: these tables can be combined or moved to the supplementary materials. Again, please specify all abbreviations used. Also, where P is less than 0.001 it is better to report as “<0.001”, rather than “0.000”.

Figs 3-6: These figures can be combined into one or two figures. This will help to quickly compare and contrast the different measures of productivity across the different treatments. Please add error bars to all. Please keep the order along the x-axis the same, both between S1 and S2 within each figure, and also across all figures. I would strongly suggest removing the black bars, which I assume (not defined anywhere) represent the combined mean for each irrigation type, across the three fertiliser treatments. There is a significant interaction factor between fertiliser and irrigation and therefore it is not appropriate to report the average across only one of those factors, because they are known to vary depending on each other. Hence, I would remove the black bar from the figures and would avoid reporting the direct comparison of the mean across all fertiliser treatments, this completely ignores the significant interaction factor and can be mis-leading (not all S1 are lower than S2) and it is not needed.

Fig. 6: change the y-axis title.

Fig. 7: This repeats data from other tables and figures and does not provide any further information. I would suggest removing this figure. Furthermore, it is not appropriate to draw a line plot where the x-axis is not continuous.

Author Response

Please see the attachment!!!

Author Response File: Author Response.pdf

Reviewer 2 Report

In this study the authors set up six different production systems under which they evaluate production parameters in cowpea. The production systems evaluated consist of three systems irrigated with waste water and driven by photovoltaic solar energy (renewable systems) and three systems irrigated with drinking water and powered by electricity from electrical grids (conventional systems). The renewable systems and the conventional systems are divided into three production systems treated with no fertilizer, organic fertilizer or mineral fertilizer. The work relies on observations of four production parameters and analyses of variance.

In general, I find the paper highly relevant and the findings promising in establishing a more sustainable cultivation of cowpea in the future. A great strength of the manuscript is its detailed description of setting up the field experiments and the irrigation system. However, I have some comments regarding the statistical design and the conclusions drawn from it.

1) A central part of the manuscript is the statistical analysis performed using Minitab. I this context, I lack a description of the model used. Judging from the tables in the manuscript, it looks like three models were applied every time analysis of a variance for a production component was performed, in that way getting three estimates of the residual variance. If that is the case, why did the authors not choose to use just one model with all relevant effects?

2) The authors could consider renaming the variance component “Energy+Water” (Table 8-11), as it gives the impression of an additive model, where the effect of water and energy are evaluated separately, which is not possible in this study, as electrical grids are always found with drinking water irrigation and photovoltaic solar energy is always found with waste water irrigation. The same goes for the interaction effect Energy*Water*Fertiliser.

3) Figure 7 shows all of the data graphed in figures 3-6. It might provide a better overview and a more concise manuscript if Figures 3-7 were combined and tables 8-11 as well.   

4) p. 2, l. 50: Change “water problem” to “water scarcity” or “water stress”.

5) p. 3, l. 90-91: “..only the central plants of each subplot being considered for data collection..” Can you state why? I assume to avoid border effects?

6) p. 4, Figure 2: Authors could colour treatments to the left according to the color they will get on the figure.

7) p. 6, l. 161: What is meant by “previously identified paper bags”?

8) p. 7, Table 4: Change legend “…in this study...” to “during the experiment” so that it matches Table 3.

9) p. 7, Table 5: Clarify in legend that the “END” data only refers to subplots treated with S2A0.

10) p. 8, l. 210-211: The correlations between the different parameters under evaluation are never presented.

11) p. 8, l. 220-222: Here the authors claim the S2A0 treatment is similar to S2A1 and S2A2 treatments, but this does not fit with the claim on p. 9, l. 236-239, where the authors state mineral fertilizer to play a “satisfactory complementary role in the production system”.

12) p. 10, Table 8: Explanations of abbreviations are missing.

13) p. 11, l. 280-281: “…it was found that the organic fertiliser (A2) was different from the mineral fertiliser (A1).” I think you are missing a “not” in this sentence.

14) p.11, Figure 3: The legend includes no information about the black bars. The same applies for Figure 4, 5 and 6.

15) p. 15, l. 377-379: “In the treatments irrigated with wastewater all the parameters under evaluation,…, had better results compared to the conventional production systems”. This is true for pairwise comparisons of systems treated with the same fertilizers only.

16) p. 15, l. 390-392. This does not appear to be supported by figures 3-6, where treatments with S2A2 seemed to outperform S2A1 or they performed similarly.

17) Figure 7: Indicate what the blue line signifies. Means or medians?

18) A key finding appears to be that wastewater irrigation can substitute for fertilizer application in cowpea production. I suggest the authors highlight this in the abstract. 

Author Response

Please see the attachment!!!

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

agronomy-575190_v2

General comments:

I would like to thank the authors for taking the time to address the previous comments. The manuscript has certainly improved and is much clearer now; however, I feel the following comments still need to be addressed. In particular, there are still several examples where conclusions are being based on non-significant differences and are therefore not supported by the results; these over-reaching conclusions should be toned down.

Broad comments:

As it is still not clear in the manuscript, I again suggest the authors make clear the important novel aspects of this research and more specifically, how it represents a significant advancement, beyond other similar studies working with wastewater in agriculture.

Specific comments:

L25-26: This was raised in the first review and has not been appropriately addressed. This sentence does not state exactly what the wastewater treatments are “better” than and in fact is contradicted by the following sentence, which states that they are not better than the conventional system with fertiliser. This sentence is therefore somewhat misleading and is not completely supported by the results. I suggest instead stating clearly what this is compared with and toning it down to what the results actually show; that the production systems with wastewater tended to show improved production compared with some of the conventional systems. More specifically, all wastewater treatments, including that with no fertiliser, showed similar yield to the conventional system with mineral fertiliser, demonstrating that treated wastewater contains sufficient nutrients to potentially replace fertilisers in the production of cowpea in northeast Brazil.

L26-29: Please re-write this sentence, the grammar needs to be improved, see suggestion above.

Introduction: It would be helpful to include a very brief summary of the experimental design and the main measurements, in the Introduction. This helps the reader to know what to anticipate and how the hypotheses/aims will be tested.

Table 2: I would suggest removing the headings “100% mineral” and “Additional”, or at least explain what they are, in the Table caption. In addition, please state clearly in the Methods section that the doses of fertilisers applied to the two irrigation types were different, as detailed in Table 2.

Table 3: This is much clearer now. The Table caption has been placed in the first cell of the table. In addition to the reference details provided for the “Reference for reuse” column, please also provide a brief explanation of what it actually is, so the reader knows what it means. Is it not possible to change the units for “Total coliforms” to allow for better values to be presented in the table? Change the definition for NS to “not specified”.

L177-179: If only a single soil sample was analysed for each system, as stated in response to my previous comment, how were they collected from the field? Was it a composite sample collected from across the field, or was it a single sample from a single location in the field. The method of collection should be specified here.

L182: Delete “And the chemical analysis:...”, also, change to “The pH was measured…”

L183: Specify here that it is “available-P (Mehlich 1)”

Table 4: Change caption to “…refers only to the subplots…”. Specify that P is ‘available-P’ measured by Mehlich-1. Please add definitions for all abbreviations, i.e. also “T”, “m”.

L224-227: Again, please be more specific, “better” than what? Same as comment L25-26.

L244-247: Again, this statement is not supported by the data. This was raised in the first review and has not appropriately addressed. There are NO differences between these two treatments for any measure of productivity, therefore you cannot state that it is higher, when it is not! In fact, for NPP, S2A1 is actually slightly lower than S2A0, which is the opposite of what is stated. Adding a sentence afterwards does not change what is said in the first sentence, the two contradict each other. Please edit this statement to accurately match the results, or remove it.

L247-251: Please revise the English spelling and grammar in this sentence, it is a difficult sentence to read.

Table 6: I would suggest either specifying the treatment abbreviations in the caption, or writing them out in the table. Each Table/Figure should be standalone.

Table 7: In the source of variation I would suggest instead stating the factor (i.e. ‘Water source’, ‘Fertiliser type’, not the levels of the factor.

L334: Delete “For the interaction” and “but with higher values”. They were not actually higher!

L389-392: Again, this statement is not supported by the data. This was raised in the first review and has not appropriately addressed. This is NOT true for NPP, NGP or W100. i.e. not all wastewater treatments were higher than S1A2 treatment, for NPP, NGP, and W100! This should be changed or removed.

Similarly, L415-417 should be revised.

L419: delete “which”

L421: specify “fertilised” systems irrigated with wastewater.

L422: delete “indicator”

Table 8: I would suggest placing the fertiliser treatments in the same order for both S2 and S1.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf