Review Reports

Round 1

Reviewer 1 Report

It appears that all of the concerns for both reviewers have been addressed adequately. There are still some awkward phrasing and typos that should be caught and corrected by careful review.

Author Response

Comments have been corrected

Author Response File: Author Response.docx

Reviewer 2 Report

The study is on the topic: Simulating the response of drought-tolerant maize 2 varieties to nitrogen application in contrasting 3 environments in the Nigeria savannas using the 4 APSIM Model. 

The authors carried out experiments vs agro-ecologies in Nigeria to determine the response of 2 maize varieties to N fertilization. The APSIM crop model was calibrated and evaluated based on experimental data. The model was applied to determine the optimum fertilizer rate for each location and maize variety. The profitability of the N rates was also assessed. 

 

I looked through the authors' responses to reviewers comment and noted that most of the comments were addressed satisfactorily. There are still some, however, that needs to be addressed.

Authors should use a more robust statistics other than the RMSE, MAE, R2 and nRMSE used to assess the performance of the model in simulating the various parameters. Example, they should use at least Willmott's index of agreement (d). The authors responded by saying "The authors agreed to use the basic statistics for the model evaluation". This is not acceptable. Authors should aim at improving the quality of their work and go beyond the "basic statistics" and include more robust statistics like Willmott's index as mentioned above.

Authors justified the use of the one-time cost of production and maize price data to estimate the profitability of maize yields for multiple years maize yield. The only way this can be acceptable is when the multiple years or seasons are regarded as replications rather than a long-term experiment as they sort to imply. Even then, there are drawbacks to this approach and this should be captured thoroughly in the discussion section. 

Below are some specific comments/suggestions to improves the quality of the paper. Careful language editing is required to improve readability, particularly the results section. The discussion sections also need to be beefed up devoid of repetition of results.

Line 251; delete "properly"

L256; delete "of the crop in the model. "

L257; replace "have" with "had" and also "before"

L260; delete "as reference varieties"

Lines 263 & 264; replace "characters" with "parameters"

L276; delete "of"

L284-285; delete "to limit the testing of the ability of the model to simulate changes in soil water 284 and nitrogen to the growing seasons"

L286; delete "were"

L 288; should read "......model simulated values using ......."

L 289; include Willmott index of agreement (d) as a measure of model performance. It is more robust than the ones you have used.

L 310; delete "model"

L 324; delete "respectively"

L339-340; should read "Simulated grain yields from the seasonal analysis across the N fertilizer rates analysed using a box and whisker plot to assess optimum N rate per site."

L345; define the letters in equation 5 and 6. What do "i"," j", "k"

L341; replace "as" with "which is"

L342; replace "to assess" with "was used to assess..."

L352; which institute?

L365; "planting" not "plating"

L374; delete "the"

L389; replace "was" with "were"

L399-390; rephrase sentence!

L399-401; delete sentence! Has no value for the calibration. Rather makes reading boring!

L402; replace "indicates the result of calibration between observed and predicted values" with shows the statistics of calibration 

L404; delete "tested"

L405-407; Extensive language editing is required for the entire manuscript.

L417; delete "with an independent experiments across N–Fertlizer applications"

L418; replace "Figure" with "Figures"

L418-422; rephrase sentence!

L429-432; You have basically reproduced the statistics on the figures in the text!

Same with L 433-436

L437-439; sentence does not make sense, rephrase it.

L451-452; should read "The response of maize to N fertilizer applications varied for both varieties at both and the locations."

L 457; "....campared to N rates of 90 ....."

L458; its not clear which CVs are for EVDT2009 and IWDC2SYNF2? 5-7% CV for 90 and 180 kg/ha for which variety?

L466-469; rephrase the sentence! Again, does it mean that each of the N rates between 90 and 180 obtained a CV of 6.5%?

L469; the basis for selecting 90 kg N/ha as optimum rate is not clear. Small CV alone can not be a basis for selecting N rate as optimum. Rather addition yield increases beyond 90 kg N are marginal!

L469-472; delete "between 90 and 180 kg N applications"

L472-474; the reason to settle on 120 N should be because "yield increases beyond 120 N is marginal. The small CV will only be an additional benefit and not the other way round!

L479; should read ".....estimated when fertilizer rates of between 90 and 180 kg N were applied."

L481-485; These 2 sentences have to be re-written for clarity. The language is bad and you are not communicating effectively

L490-493; delete the information on the CVs

L496; should read ".... across N application rates ..." not " ....... across N application rate"

L505-507; High compared to what? What about kano who has the least probability of obtaining yields more than 4000 for the IWDC2? Rephrase your sentence.

L521-522, L523-524; should read "....all the years for application rates between 90 and 180 kg N ha-1.

L527; delete "only"

L532; replace "would be" with "was"

L533; delete "would be expected"

L549-551; what do you mean by "...... resulting to labour loss for farmers"?

 

Discussions

L560; delete "(≤ 12%)" and "(R2 ≥ 0.8) "

L575-580; don't repeat results in this section. You can refer to it but not listing all the values earlier reported in the results section.

L582-584; How do you know that the differences were significant or not? I didn't see any statistical analysis mentioned in the methodology that will enable you arrive at this conclusion. If you did some statistics, provide it in the methodology section!

L618-628; The recommended N rate per your study varies based on maize variety and agro-ecology. Similarly, it varied from the other two studies in Ghana you referred to. So, what does that mean to fertilizer recommendations? '

L636; delete "(2009) "

Author Response

Comments have been corrected 

Author Response File: Author Response.docx

Reviewer 3 Report

Thanks to numerous varieties, maize copes well in various climatic conditions. It has a very large capacity, but requires a lot of heat, but not excessive moisture. It is one of the most widely grown grain crops in Nigeria and an important plant for human consumption.

The savannah soils in Nigeria are mostly sandy liqueurs, acrysoles and cambisols with low clay activity (such as kaolinite), low organic matter content, low nutrient reserves and susceptible to water and wind erosion. This state of affairs limits the intensification of maize production. The use of nitrogen fertilizers is cost and has a low recovery rate due to losses from denitrification, ammonia volatilization, runoff and leaching.

Therefore, studies simulating the response of drought tolerant maize varieties to nitrogen application in the contrasting environments of the Nigerian savannas are useful especially with presentation of  profitability scenarios analysis.

 Thus the topic of this paper is of paramount importance when considering the fact that studies are essential to understand the effect of different management practices on soil and plant system. In addition the topic is closely falls within the aims and scope of the journal. Material and methods need some modifications. The data provided are sufficient and the statistical analysis of the results is very well presented. Figures clearly present the data. The discussion of results focus on the main points while justification of the findings are well supported by references.

 

In material and methods the authors presented 4 sites : were Abuja (9° 9′N 7° 20′E 447 m a.s.l.), in the southern Guinea savanna (SGS) zone, Zaria (11° 1′N 7° 37′E 681 m a.s.l.) in the northern Guinea savanna (NGS) zone and Kano  (11°59′N 8°25′E 466 m a.s.l.) and Dambatta (12°19′N 8° 31′E 504 m a.s.l.) in the Sudan Savanna (SS) zone. The results are presnted for Abuja, Zaria, Kano.

 

The descriptions in Figures 1, 2 and 6 are inaccurate. In Figure 1, a-f have the letters a-d. The author could have given the names of the places above Fig. 6. In fig. 2 d there is probably a typing error (observed is used twice). Figures 6 have the same diagram a as b.

Author Response

Comments have been corrected.

Author Response File: Author Response.docx

Reviewer 4 Report

To whom it may concern,

Thanks for the invitation. I have gone through the manuscript titled ‘Simulating the response of drought–tolerant maize varieties to nitrogen application in contrasting environments in the Nigeria savannas using the APSIM Model’. Here are my comments:

Abstract

Here the author mainly focused on what have been done and the results of this study. However, the reason why the author conduct this work, the aim and the significance of this study was missing.

The full name of ‘APSIM’ should be provided.

Introduction

Line 49,’Over’ should not be bold

Line 50, World should be replaced by world

The full name of ‘CERES’ should be provided.

Line 82, the full name and abbreviation of ‘APSIM’ has been mentioned. So in Line 90, only abbreviation is needed. Check the other abbreviations.

Be consistent to apply APSIM model or module.

The hypothesis and significance of this work need to be added in the end of introduction.

M and M

Line 166, Three (3)? Line 188, five (5) levels? The numbers in bracket is not necessary. The same for Line 243, Line 319, Line 374

Line 172, (15m²) should be (15 m²); 0.75m should be 0.75 m.

Line 189, 30kgha^–1. The author need to carefully check these small problems throughout the whole manuscript.

Line 197, 5WAP / 5 WAP

Line 206, so what the author mean by R2 stage?

Results, discussion and conclusion

Figure 1: The color of mean rainfall need to be adjust to the same; Figure 1d was not mentioned in the main text; fiugre1a-f, so where is the figure 1e and figure 1f. Or the figure caption is not clear so that it is easy to raise misunderstand.

When the discussion was made, it is recommended that the author cited the results (e.g. which table or figure) to facilitate the reader to follow and check. For instance, the author cited the results section of figure 5 in Line 631.

The conclusion should focus on the main results and conclusion as well as the significance of this study instead of only the results in detail.

Author Response

Comments have been corrected.

Author Response File: Author Response.docx

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

The study entitled:

Simulating the response of drought-tolerant maize varieties to nitrogen application in contrasting environments in the Nigeria savannas using the APSIM model.

Authors conducted experiments in 2017 & 2018 at different locations in Northern Guinea Savanna, Southern Guinea Savanna and Soudan Savanna zones. Part of the data collected under limited nutrient and water stress to calibrate 2 maturity duration maize varieties. Another set of data was used to evaluate the performance of the model. The model was subsequently used to run a seasonal analysis over 31 years of weather data to simulate maize yield under different N levels and used the simulated yields to determine optimum site-specific fertilizer recommendation. 

The subject matter of relevance especially in Sub Sahara Africa where efficient use of inorganic fertilizer is a challenge.

Generally, the paper is well written but needs some grammar checks and several issues that have to be addressed. In the materials and methods, no mention was made of statistical analysis. However, several references were made to significant differences.

The procedure for calculating the profitability of N fertilizer use was not clear enough. Why use price and cost of production data for the experiment period to calculate the profitability of simulated yields spanning the period 1985-2015? Also, provide a Table showing the various cost elements and unit cost as well as price data.

The literature cited in the test was not in the list of references. And it appears the list of references do not match the subject area. This made it impossible to cross-check some of the statements attributed to some authors.

 

The discussion section needs to be improved significantly after issues with interpreting the exceedance graphs and more objective analysis is used to determine the optimum net income under the scenario analysis section.

This will then flow into the conclusions

 

These issues and the detailed ones below need to be addressed before decision made on accepting the paper.

Abstract

Lines 24-26: Rephrase sentence for clarity

What is the policy implications of your findings? Provide one concluding sentence on that.

 

Introduction

Apart from indicating the objective of the story which I think is important, the justification for the study is not strong enough. The authors should do a better job at enhancing the justification for the study. There are a few editorial issues too to address:

Line 45: What is the current status of maize? This sentence doesn't flow properly with the earlier one.

Line 70: Replace "difficult" with "unrealistic"

Line 75: Replace "according needs" with "according to the needs.."

Line 77: What do you mean by "slow"?

Line 87: Replace "also help with" with "be used as a tool to support"

Line 99: Replace "system" with "systems"

Line 107: Fosu-Mensah et al not in the list of references

Provide a concise hypothesis for your study.

 

Materials & Methods

Line 141: what do you mean by profile horizon?

Line 139: Add “s” to “site” and delete “and”

Lines 143-145: what parameters were analysed? State them and provide better information on equipment used in the analysis of the various parameters? Which measured parameters served as input to the pedo-transfer function?

Line 143: Delete “analyses”

Line 154: very close is subjective! Provide a figure/value in km

Line 156: What does NIMET stand for? Acronyms are supposed to be explained the first time they are mention in a text!

Line 157: Add “s” to “soil”

Line 179: replace "A trial" with "Another trial"

Line 190: "...Muriate of Potash" or "...Moring of Potash.."?

Line 191: What quantity of nitrogen did you apply?

Line 202: How will you justify using data from 1m2 in estimating the yield of Maize? Will 1m2 capture the diversity in each plot?

Line 206: how do measure and calculate the number of ears and number of grains? How did you capture the number of ears into the model?

Line 234: replace "model" with "module"

Line 235: replace "inputs" with "input"

Lines 240-241: List the variables that were obtained from the literature.

Line 244: what do you mean by "properly setup"?

Line 247: not in the list of reference. Different unrelated literature is assigned "43". By the way, are you sure Willmott has authored work on thermal time calculations?

Line 248: why set the optimum temperature at 30?

Line 258: mention the data that was measured and those that were estimated

Line 259: replace "an" with "the"

Line 268: Same reference issue here!

Lines 275-277: Rewrite this sentence for clarity. Do you mean to say that the soil and residue variables were reset at the beginning of each simulation to limit the testing of the ability of the model to simulate changes in soil water and nitrogen to the growing seasons? Imply no carry-over effect to the following season? Please clarify

Line 279: insert a comma after "maturity,"

Lines 281-283: Why not include any of the more robust statistical measures of model performance such as Willmott's d index and coefficient of model efficiency?

Line 293: Which is providing the percentage error? RMSE? How is the unit of RMSE percentage?

Lines 295-296: Citation (Jamieson et al., not in the list of references.

Line 300: Replace "long term" with "seasonal"

Lines 313-315: This sowing rule is different from the one stated in lines 308-309. So the question is which one of the two was used?

Table 1: Provide data on finert, fbiom values used for the various soils?

Calculation of profitability:

You simulated yields over 31 years and used to determine profitability. But the used cost of production and stalk & grain price data for the experimental year to determine profitability instead of the price and cost of production values of the corresponding years to determine the net revenue? These data vary over time and CPI will also vary year on year. So what is the justification for the approach you used?

 

Provide a table on the data on input costs as well as prices of stalk and grain used for the determination of profitability.

Line 342: replace "in" with "on"

Line 343: Replace "under different" with "under each of the different"

Line 345: How was the value of the stalk obtained?

 

Results

Lines 353-373: Move this section to Materials and methods and avoid list what is already obvious from the Table.

Line 378: delete "were recorded"

Line 382: replace "variability in rainfall" with "differences in total rainfall amounts"

Lines 397-398: Check the sentence again for completeness! ".....and >10% respectively."

Line 400: should be "....... RMSEn 6%."

Lines 407-408: What about grain and total matter yields?

Line 409: I don't see any "N" in the table. Or you mean to state: Number of observations used for estimating the statistics is 9? State it more appropriately

 

Lines 420: round up the decimals to the nearest whole numbers

 

Figures 2 & 3

Calculation of the r2 should be redone.

The r2 calculations should be based on the regression line through the origin (0) and not with an intercept. Remember you are comparing observe to simulated.

 

Also, keep the RMSEn as whole numbers

Line 440: Replace "long-term" with "seasonal"

Text inline 440 to 444 more appropriate as a footnote to figures 4, 5 & 6. Delete from the text.

Line 441: Replace "over the long-term" with "for 31 years

 

Line 444: How did you arrive at the significant effect without any statistical analysis to that effect mentioned in the materials and methods section?

 

Line 447: delete "(50% quartile)"

 

Lines 449-450: The differences in variability you mention are difficult to visualize. I wonder if these differences will be significant to worth mentioning. Provide CV values.

 

Assuming this is true, what will be the reason for this? The reverse is normally reported in the literature

Figure 4-6: Explanations on the box plots should be provided as a footnote to the figure caption.

The text in lines 440 to 444 should be moved here to serve as a footnote

 

Line 457-458: Are these values median or mean yields for the various N levels? Please specify

 

Lines 460-462: What do you mean by "...with stable grain yield ranging from 3792-4668 kg ha-1 at 90 kg N." What is the stability here referring to? Is the range referred to here min to max yields under 90 N or 25th percentile to 75th percentile? Please clarify to reduce the ambiguity in your sentence.

 

Line 463: Replace "no significant" with "marginal"

Line 469: ".......the simulated yields...." Are these median or mean simulated yields? Please specify

 

Line 471 and many other places, you kept referring to low or high variability. The question here is; what is your measure of variability? What values do you classify as high and what values are classified as low? Your materials and methods section should provide how you measured or assessed variability.

 

Line 473-475: Rephrase the sentence for better clarity. What do you mean by stable yield? What was the measure used to measure stability? Please provide this in the materials and methods section.

 

Figure 4-6: provide a unit for the x-axis. I suggest you combine the 3 figures into one as you did for figures. Explanations on the box plots should be provided as a footnote to the figure caption.

The text in lines 440 to 444 should be moved here to serve as a footnote

Line 486: How did you arrive at "the desired yield of 3000 kg/ha"?

Line 487: replace "almost 95%" with "all". Check the figure again!

 

Interpretation of exceedance graph (lines 488-492) is not very clear.

Can you explain why you used 3000 kg/ha for 2009EVDT and 4000 kg/ha for IWD? Again all the yields obtained from N levels above 90 N exceeds 4000 kg/ha. So why indicate 95% of the years instead of 100%? explain to the reader!

Line 505: Why 95% of years when the yields under the 90-180N were all higher than 3000 kg/ha?

Line 507-508: 95% of the years is not correct (per the figure 90-180N. Please check figure again!

 

Lines 508-509: Why compare these two when the threshold you measure each with is different? Rephrase the sentence to clarify that "respective threshold" ?

Lines 534-535: Is labour the only cost element when no fertilizer was applied?

 

Figure 8: What is the label for the x-axis?

Line 539: I thought it should have been "scenario analysis" instead of "risk analysis"?

Line 541: Provide a reference for the exchange rate used.

Line 571: Replace "limitation" with "limiting"

Lines 581-583: Rephrase sentence

Lines 595-597: Rephrase sentence!

 

Lines 597-604:

Rephrase: I think the yield differences between the two varieties are more obvious in the GS than the SS. The lower yield difference in the SS could be explained by the shorter length of the growing season, while the longer length of the growing season in GS supports the longer maturity variety with higher yield potential.

 

Rephrase your sentences for better clarity Lines 597-604

 

Lines 645-647: Is the difference between the net income from 90 and 120 N big enough to choose 120N over 90N? You will need some form of statistics to arrive at a decision which is not subjective as it appears in this case.

Reviewer 2 Report

This research article demonstrates the utility of the APSIM-maize growth model for predicting yield harvest on two drought-tolerant maize varieties grown in Nigerian savannas and reports results for optimal nitrogen applications for three regions. A major strength lies in the proven accuracy and adaptability of the APSIM model, as demonstrated by its employment in numerous other studies. The results presented are not surprising but are potentially impactful for Nigerian farmers and government policy advisors in the region. Crop modeling showed that the maize varieties tested here (2009EVDT and IWDC2SYNF2) should be fertilized with between 90- and 120-kg N per hectare, amounts in keeping with other recommendations for optimal N application for maize. The paper also presents economic analyses, which will be useful in helping farmers conduct cost-benefit analyses for their particular situations.

The paper is very well-written and thorough in describing all aspects of the methodology, results, and interpretations. Some detail could be omitted or moved to supplemental, as the paper is long and may be improved with conciseness. I’ve suggested a few places that could be condensed below, but generally the entire paper could be condensed a little.

I do not have any major concerns with the paper, but minor comments follow:

Line 92: Expand on why the APSIM model is more suitable for African farming systems, or remove than probably unneeded claim.

Line 94 through 116: Major condensation could be done here. Detailed results from other studies are not needed. Stating that the model has been effective and citing the papers is good enough unless there is a particular finding that is immediately relevant to the current work.

Only a single soil sample was analyzed from each site? Is there information regarding the soil content variation within these sites? Do the findings align with other published soil analyses done on these sites? I think it would have been better to take and analyze four separate samples spread apart within each site.

Table 2 should be moved to Supplemental.

Line 526: A lot of this detail is apparent from the figures and could be omitted from the text. Same is true for some of the other results presented in graphs.

Line 560 (and elsewhere): Be careful to not repeat results in the discussion, only if necessary to make a particular point in discussion.

Reviewer 3 Report

The present study entitled 'Simulating the response of drought-tolerant maize varieties to nitrogen application in contrasting environments in the Nigeria savannas using the APSIM Model' is based on APSIM model with input data from two years of field study at three locations with two maize varieties. The methods used are very well described and results are well explained. 

However, there are some basic questions about the study:

1. Information regarding the area and adaptability of the two chosen varieties is missing. Normally in case of maize many hybrids are recommended by different institutes or companies in a region and they persist for 2-3 years and are replaced with new hybrids having better characteristics after that. In this study a detailed simulation has been done using these two varieties and nitrogen dose has been recommended for these two for three regions of Nigeria. However, it may be possible that these varieties may not exist after 1-2 years and the whole exercise may become futile. As the authors have mentioned that among different regions of Nigerian savannas, there is wide variation in soil characteristics and weather parameters so why not recommend fertilizer dose based on soil and weather type. Soil fertility status of different regions may be used and general classification of varieties (low, medium and high yielding) may be used to recommend fertilizer nitrogen rate for different regions rather than giving recommendation for specific varieties. 
2. The authors have used 30 years of weather data for long term simulation, however, again these varieties may not exist for a longer duration. 
3. There are few minor mistakes like at L-45 to 47 font size is different. There are few spelling mistakes and some missing information like at L-356 what is <4.0 g per kg?