Landscape and Micronutrient Fertilizer Effect on Agro-Fortified Wheat and Teff Grain Nutrient Concentration in Western Amhara

Round 1

Reviewer 1 Report

The article sets an interesting goal.
However, I believe all the soil analyses, and weather conditions recorded during the experimental periods, are needed in order to understand specifically all the interactions that may have occurred during the speriemental trials.
The materials and methods sections are insufficiently written, sound unscientific, and are in my opinion too long.
The results section needs in my opinion to be increased and improved, if possible reporting other findings, as the target of the journal is high and related to agronomy, moreover, the graphs are very poor.
The discussion section is in my opinion completely lacking sufficient explanations to clarify the readers' ideas, also very superficial.

Ultimately, in my opinion, the article should be rejected as it is not properly structured

Author Response

Responses to Reviewer 1’s comments

The article sets an interesting goal.

However, I believe all the soil analyses, and weather conditions recorded during the experimental periods, are needed in order to understand specifically all the interactions that may have occurred during the speriemental trials.

We value the reviewer’s suggestions. We provide background soil analyses data to support our findings. Rainfall data is now included as Figure 2 and also cited in the manuscript [Lines 154 – 156].

The materials and methods sections are insufficiently written, sound unscientific, and are in my opinion too long.

We provided detailed information to reflect changes in treatments between the two-cropping season. We revised specific sections in the Methods section as suggested by Reviewer 2. We are open to work with the reviewer on specific sections they suggest for us to revise.

The results section needs in my opinion to be increased and improved, if possible reporting other findings, as the target of the journal is high and related to agronomy, moreover, the graphs are very poor.

All the graphs are of good statistical quality.  All origins are shown, even when an axis has to be truncated to do this, there are error bars on all estimates, and the caption specifies exactly what these are.  Point estimates are not presented as bar charts.  All axes are clearly labelled using Royal Society standard practice for units.  The referee raises no specific objections, so we have made no edits.  If the editor has any particular concerns which he or she would like to raise about the graphs then we will be glad to make necessary changes.

The discussion section is in my opinion completely lacking sufficient explanations to clarify the readers' ideas, also very superficial.

We disagree with the Reviewer’s comment on our points being superficial. We supported our discussion points with earlier published work, some of which was published by the authors of this work (i.e. Desta et al on Zn adsorption and desorption as influenced by different soil physico-chemical properties along a landscape gradient and Manzeke et al 2020 on increased grain Zn concentration under low N compared to high N rates). We are open to work with the Reviewer on specific points which they deem superficial.

Ultimately, in my opinion, the article should be rejected as it is not properly structured

Authors made substantial improvement to the manuscript based on specific comments raised by the other Reviewer. We are open to work with the Reviewer on areas of the manuscript they deem improperly structured.

Reviewer 2 Report

Manuscript has a great potential. There is great inconsistency in experimental design, regarding treatments across seasons, with no data on soil type and meteorological conditions, so efficiency of applied treatments is questionable. In general, statistics underlined important points, but repeatability and given conclusion could not be supported by achieved results.

There are also writing errors, so manuscript should be checked, accordingly.

Line 83: to crops.

Table 1: In the Table title there are Test crops, but they are not present in table, please, correct.

Line 141-151: How many farms were included in the experiment? Please, introduce plot dimensions, from the smallest to the largest one.

What is about soil analysis? How was recommendation for fertilizer application given, based on the soil analysis, or amounts mainly used for these crops?

Line 167-182: Please include Se addition, accordingly to the Table 2, to be clear about applied treatments.

Table 3: Does it mean that teff was sown on nitisoils in 2018 and on vertisoils in 2019? What is about wheat (Table 2), there is no notice in the text or in the Table that wheat was sown on the same or different soil types.

What is about grain yield? Cost-effectiveness of any biofortification method is questionable if reasonable yield was not achieved. Providing the greater amounts of micro-nutrients could support crop growth and thus increase yield, as a result, but in some cases, when yield is low (due to the impact of agro-meteorological factors, such as drought) relatively high concentration of micro-nutrients in grain is cost inefficient. So it is recommendable to include yield of both crops and correlate them with results achieved in Zn and Se concentration in grain.

Line 264-283: Why multi-elemental analysis was described, since only Zn and Se concentration in grain was present in manuscript?

Line 306-308: You meant, when C1 and C3 were compared?

There are no titles for Figures and Tables in Supplementary material, so the reader could only guess which results from Supplementary material were described in Results.

Line 436-442: It seems that in this paragraph Se content was discussed, so this part of the text should be removed to the previous paragraph (for instance, after line 429).

Line 455-470: This part of the Discussion is very important, so if data on soil composition could be imported, then full meaning of achieved results could be supported.

Line 472-479: Of course, genotypes with high adaptability to various conditions, particularly to stressful conditions and low-inputs are always highly efficient, but mainly with lower yield potential.

Author Response

Responses to Reviewer 2’s comments

Manuscript has a great potential. There is great inconsistency in experimental design, regarding treatments across seasons, with no data on soil type and meteorological conditions, so efficiency of applied treatments is questionable. In general, statistics underlined important points, but repeatability and given conclusion could not be supported by achieved results.

We value the Reviewer’s comments on experimental design, treatments and lack of soil analytical data. The experimental design was structured in randomized complete block design, replicated across farms with farm fields selected at random along a landscape gradient: foot slope, mid-slope, and hillslope. This was to capture effects of landscape on grain micronutrient concentration to address complexities in deployment of micronutrient fertilizers in heterogenous smallholder farming systems. Due to unforeseen leaf scotching from co-application of foliar Zn and Se, treatment 3 and 4 was changed in season 2 and this was explicitly elaborated in the paper. We provided soil analytical data published by one of the co-authors to support our discussion. We however disagree with the reviewer that our conclusions could not be supported by achieved results. We are open to work with the reviewer on ares of the conclusion they feel are not supported by our findings.

There are also writing errors, so manuscript should be checked, accordingly.

We checked the manuscript for typo errors.

Line 83: to crops.

Revised as suggested by the Reviewer.

Table 1: In the Table title there are Test crops, but they are not present in table, please, correct.

We acknowledge the oversight. Revised Table 1 caption as suggested by the Reviewer.

Line 141-151: How many farms were included in the experiment? Please, introduce plot dimensions, from the smallest to the largest one.

The number of farmers has now been mentioned in the Methods section of the paper [Lines 176 to 178]. The size of the treatment plot has also now been mentioned as “Each treatment plot had 5m*5m dimensions”.

What is about soil analysis? How was recommendation for fertilizer application given, based on the soil analysis, or amounts mainly used for these crops?

Fertilizer application rates were applied based on soil type and crop type. For example, wheat grown on Nitisols and Vertisols received similar N, P, S, K and B rates of 138 kg N ha^-1, 92 kg P₂O₅ ha^-1, 16.96 kg SO₃ ha^-1, 40 K₂O ha^-1 and 0.24 kg B ha^-1. Teff grown on Nitisols received 40 kg N ha^-1, 60 kg P₂O₅ ha^-1, 11.06 kg SO₃ ha^-1, 40 K₂O ha^-1 and 0.16 kg B ha^-1 while teff grown on Vertisols received larger rates of 80 kg N ha^-1, 46 kg P₂O₅ ha^-1, 8.48 kg SO₃ ha^-1, 40 K₂O ha^-1 and 0.12 kg B ha^-1 due to increased leaching and volatilization on these soils [Lines 220 to 232].

Line 167-182: Please include Se addition, accordingly to the Table 2, to be clear about applied treatments.

Revised as suggested by the Reviewer [Lines 210-212].

Table 3: Does it mean that teff was sown on nitisoils in 2018 and on vertisoils in 2019? What is about wheat (Table 2), there is no notice in the text or in the Table that wheat was sown on the same or different soil types.

Authors acknowledge the lack of clarity in Table 3 (now Table 4) and revised Table 3 caption to reflect soil types and Table 4 to specify soils and treatments for teff in 2018 and 2019 season.

What is about grain yield? Cost-effectiveness of any biofortification method is questionable if reasonable yield was not achieved. Providing the greater amounts of micro-nutrients could support crop growth and thus increase yield, as a result, but in some cases, when yield is low (due to the impact of agro-meteorological factors, such as drought) relatively high concentration of micro-nutrients in grain is cost inefficient. So it is recommendable to include yield of both crops and correlate them with results achieved in Zn and Se concentration in grain.

 

It is certainly true that reduced yield (due to environmental factors such as those which the reviewer mentioned) could lead to increased micronutrient concentration in grain in some particular case.  However, because we have used a randomized complete block design, with farms as blocks, and treatments randomized among the plots on each block, these environmental factors can be treated as random block and residual effects.  That is to say, they are as likely to increase Zn concentration in a control plot as in a plot with Zn applied.  The treatment means that we estimate are therefore design-unbiased for such effects, and we can state with confidence that, over the range of conditions encountered, there is a mean benefit from the intervention.  Now some of the residual variation or block effects might be accounted for if yield was included as a covariate, but that has no bearing on our conclusion about the treatments, it simply shuffles some of the random variation between different rows of the ANOVA table.  In short, because the practical objective of interest is to understand impacts on micronutrient concentrations in grain (and hence on micronutrient intake from a standard portion of grain), our analysis is the correct one to use.

Line 264-283: Why multi-elemental analysis was described, since only Zn and Se concentration in grain was present in manuscript?

Revised as suggested by the reviewer to only focus on Zn and Se [Lines 309 and 322].

Line 306-308: You meant, when C1 and C3 were compared?

This sentence has been revised for clarity to “Strong evidence of fertilizer application effects on grain Zn concentration in wheat was evident in both C1 (No Zn input vs Zn) and C2 (Soil vs Soil + Foliar application); (P<0.001) (ANOVA Table 1 Supplementary file)”. [Lines 351-353].

There are no titles for Figures and Tables in Supplementary material, so the reader could only guess which results from Supplementary material were described in Results.

We acknowledge this valuable comment from the Reviewer. Captions have now been provided to link main text and supplementary material.

Line 436-442: It seems that in this paragraph Se content was discussed, so this part of the text should be removed to the previous paragraph (for instance, after line 429).

Discussion section revised by authors for clarity.

Line 455-470: This part of the Discussion is very important, so if data on soil composition could be imported, then full meaning of achieved results could be supported.

We have now provided soils data and acknowledged source of data (Desta et al 20210.

Line 472-479: Of course, genotypes with high adaptability to various conditions, particularly to stressful conditions and low-inputs are always highly efficient, but mainly with lower yield potential.

We note this important point and revised the sentence to denote yield reductions from resilient crops.

Round 2

Reviewer 1 Report

The manuscript is now well presented and complete, the graphs remaining very poor in my opinion, but is only opinion.

Reviewer 2 Report

Manuscript is considerably improved. Some suggestion still be considered to improve its quality.

Data are missing in Figure 2.

Table 2: I’m guessing that Se analysis in soil samples was not done? This is the first time I heard abbot Aqua regia used for N determination. There is great amount of clay in soil samples and also, they are acidic, particularly soil from Markuma location, so it is questionable to use Olsen method for P determination. According to these results, it is not recommendable to use urea for fertilization, since it increases soil acidity and thus could affect crop growth.

Still, grain yield is missing to additionally support effectiveness of applied fertilization combinations, due to the fact that optimal nutrition, particularly when micronutrients were considered, provides greater yield potential. On the other hand, highly efficient genotypes are not always those with greater yield potential.

Line 514-538: This is why inclusion of soil composition is important, since topography partially impact soil mechanical and chemical composition, as well as pH. On the other hand, soil per se, including maternal rock and SOM has huge influence on micronutrient's availability. It is well known that some nutrients, particularly N (and many others on sandy, light soils) migrate over profile, as well as from higher to lower slopes, accumulating at lower slopes, making them more fertile. You used perfect references in this part of the text, and you may discuss deeper own results, using them, such as pH, SOM and clay, which mainly interfere micronutrients availability and absorption.

Line 563-569: Maybe you could to consider some additional comment, such as: landscape position could be one of the first factors to consider, when some agricultural production is organizing, particularly when bio-fortification programs are implementing.