Weed Control and Selectivity of Four Herbicides Applied in Pre-Emergence on Two Sunflower Cultivars
Round 1
Reviewer 1 Report
Although the study has merit, the way in which it was executed, the lack of information provided in the material and methods, the unclear or incorrectly relaying of results in all tables as well as the lack of repetition across seasons unfortunately results in this manuscript not being suitable for publication. I found the document difficult to follow, because of the issues raised above and, as I’m not sure of how analyses were conducted, I cannot provide any useful comments on the discussion and conclusion. Below are some of the concerns/comments I had.
1) Grammar below standard – requires editing
2) New title suggested: ‘Weed control and selectivity of pre-emergence herbicides on sunflower’
Materials and methods
1) Where both trials planted on 15 March?
2) Were both trials sprayed on the 15th of March as well?
3) What type of soil cultivation practice was followed… conventional or conservational?
4) Why were the same dosage rates applied at the two soil types? Generally greater concentrations are required with clay soils, or then lower concentrations at sandy soils. Applying the same rates and then comparing them directly is not suitable for a weed control publication. Please indicate the reasoning behind this approach.
5) Line 81 “The experimental design used was a randomized block design with a 2 x 6 x 2 factorial scheme with four repetitions” If two different sites were utilized (only clay soil and one sandy soil), I can not see how you can analyses them together as a randomized block design, as the soil type was not also randomized (assuming that the treatments were indeed randomised at each of the two localities i.e. different trial plans). The general approach in this case that one must first test if the error variations were homogenous or of similar size. As there are only two experiments, an F-test should have been used to test the error variances. If the variances were heterogenous, weighted Anovas should have been used when the two experiments were analysed. The problem is however that the tables presented in the document indicate that each cultivar was analysed separately per soil type. Hence the M&M must be adjusted to indicate this, as well as why this approach was followed.
6) Did sufficient rainfall fell within 7-10 days after application? If sufficient rainfall did not fall, then it is insufficient to just publish one year’s data… repetition across seasons will be required.
7) What area per plot was used for screening/evaluations? The use of the term “useful area” is not a suitable reference for a scientific publication.
Results
1) Line 25-26: “No differences were observed for the level of injury to plants in all the seasons evaluated, and the results were similar to those of the control (no injury)”. The same dose rates were used for sand and clay soils. If the normal recommended dose rates for clay and soil were not applied, how do you know that you tested at the level producers will apply it…?
2) The relevance of Figure 1 is debatable. It does not provide an accurate indication of what happed at the respective two sites, or give clarity as to when the rainfall fell directly after planting, The figure layout is also problematic in what appear to be incompatible fonts being used on the x-axis.
3) Tables: the “a” and “b”s allocated to all tables cannot be correct if the two locality x 2 cultivar x 6 treatments were analysed in one ANOVA. How were these data sets analysed…? It does not appear to be as the “The experimental design used was a randomized block design with a 2 x 6 x 2 factorial scheme with four repetitions” indicated? (line 81). All the tables appear to have been created by evaluating the cultivars separately per locality… i.e. four different ANOVAs as each has cultivar x locality combination has it’s own CV values and statistical differences (“a”s and “b”s).
Discussion and conclusion were not reviewed questions regarding Materials & Methods must fist be clarified before an evaluation of the results can be done.
Comments for author File: 
Comments.pdf
Author Response
Reviewer #1:
Although the study has merit, the way in which it was executed, the lack of information provided in the material and methods, the unclear or incorrectly relaying of results in all tables as well as the lack of repetition across seasons unfortunately results in this manuscript not being suitable for publication. I found the document difficult to follow, because of the issues raised above and, as I’m not sure of how analyses were conducted, I cannot provide any useful comments on the discussion and conclusion. Below are some of the concerns/comments I had.
The entire manuscript has been revised for the reader's better understanding.
1) Grammar below standard – requires editing
2) New title suggested: ‘Weed control and selectivity of pre-emergence herbicides on sunflower’
The suggestions have been inserted into the manuscript.
Materials and methods
1) Where both trials planted on 15 March?
Yes, in both areas.
2) Were both trials sprayed on the 15th of March as well?
Yes, the herbicides were applied on the same day as sowing.
3) What type of soil cultivation practice was followed… conventional or conservational?
Conventional system.
4) Why were the same dosage rates applied at the two soil types? Generally greater concentrations are required with clay soils, or then lower concentrations at sandy soils. Applying the same rates and then comparing them directly is not suitable for a weed control publication. Please indicate the reasoning behind this approach.
Yes, in both soil types the same dose of herbicides (highest field recommended dose in the label) was applied. In Brazil, for sunflower cultivation many residual herbicides are recommended only one dose, with no difference between soil types. We agree, and the labels of Brazilian herbicides need to be updated.
5) Line 81 “The experimental design used was a randomized block design with a 2 x 6 x 2 factorial scheme with four repetitions” If two different sites were utilized (only clay soil and one sandy soil), I can not see how you can analyses them together as a randomized block design, as the soil type was not also randomized (assuming that the treatments were indeed randomised at each of the two localities i.e. different trial plans). The general approach in this case that one must first test if the error variations were homogenous or of similar size. As there are only two experiments, an F-test should have been used to test the error variances. If the variances were heterogenous, weighted Anovas should have been used when the two experiments were analysed. The problem is however that the tables presented in the document indicate that each cultivar was analysed separately per soil type. Hence the M&M must be adjusted to indicate this, as well as why this approach was followed.
This has been corrected.
The experimental design used was a randomized block design with a 2 x 6 factorial scheme with four repetitions, analyzing two cultivars (M 734 and CF 101) and six control methods (four chemicals and two controls, one without application and the other with weeding), in two soils with different textures (clayey and sandy), with independent experiments.
6) Did sufficient rainfall fell within 7-10 days after application? If sufficient rainfall did not fall, then it is insufficient to just publish one year’s data… repetition across seasons will be required.
Yes, had enough rain.
7) What area per plot was used for screening/evaluations? The use of the term “useful area” is not a suitable reference for a scientific publication.
Sampled area (middle two rows).
Results
1) Line 25-26: “No differences were observed for the level of injury to plants in all the seasons evaluated, and the results were similar to those of the control (no injury)”. The same dose rates were used for sand and clay soils. If the normal recommended dose rates for clay and soil were not applied, how do you know that you tested at the level producers will apply it…?
Yes, in both soil types the same dose of herbicides (highest field recommended dose in the label) was applied. In Brazil, for sunflower cultivation many residual herbicides are recommended only one dose, with no difference between soil types. We agree, and the labels of Brazilian herbicides need to be updated.
2) The relevance of Figure 1 is debatable. It does not provide an accurate indication of what happed at the respective two sites, or give clarity as to when the rainfall fell directly after planting, The figure layout is also problematic in what appear to be incompatible fonts being used on the x-axis.
Figure 1 was unformatted and has been corrected.
3) Tables: the “a” and “b”s allocated to all tables cannot be correct if the two locality x 2 cultivar x 6 treatments were analysed in one ANOVA. How were these data sets analysed…? It does not appear to be as the “The experimental design used was a randomized block design with a 2 x 6 x 2 factorial scheme with four repetitions” indicated? (line 81). All the tables appear to have been created by evaluating the cultivars separately per locality… i.e. four different ANOVAs as each has cultivar x locality combination has it’s own CV values and statistical differences (“a”s and “b”s).
The experimental design used was a randomized block design with a 2 x 6 factorial scheme with four repetitions, analyzing two cultivars (M 734 and CF 101) and six control methods (four chemicals and two controls, one without application and the other with weeding), in two soils with different textures (clayey and sandy), with independent experiments.
Discussion and conclusion were not reviewed questions regarding Materials & Methods must fist be clarified before an evaluation of the results can be done.
Okay, we are in agreement.
Reviewer 2 Report
The major flaw of this manuscript is poor data analysis. There are 3 factors in this study. It is not clear whether there is an interaction between them although it is stated in the methodology that ANOVA was carried out to determine the interaction. Since it is not clear about the interaction effects, data presentation, data interpretation, conclusion and abstract are greatly affected.
Author Response
Reviewer #2:
The major flaw of this manuscript is poor data analysis. There are 3 factors in this study. It is not clear whether there is an interaction between them although it is stated in the methodology that ANOVA was carried out to determine the interaction. Since it is not clear about the interaction effects, data presentation, data interpretation, conclusion and abstract are greatly affected.
The experimental design used was a randomized block design with a 2 x 6 factorial scheme with four repetitions, analyzing two cultivars (M 734 and CF 101) and six control methods (four chemicals and two controls, one without application and the other with weeding), in two soils with different textures (clayey and sandy), with independent experiments.
Reviewer 3 Report
The result of the article titled “Weed control and selectivity of herbicide applied in pre-emergence with a backpack sprayer on sunflower ” is interesting.
To improve the manuscript It would be better if the title were revised. The tile does not need to include backpack sprayers, for example.
The introduction should be improved. They should compare the pre-emergence herbicide to the post-emergence herbicide for sunflower, explaining why each is better in the location. The authors need to explain why these S-metolachlor, sulfentrazone , flumioxazin, and trifluralin herbicides were chosen and explain what is the initial phase.
Materials and methods: The author should explain when herbicide was applied (how many days after or before planting), the method of injury evaluation (crop? weed? visual), and yield harvesting. I would recommend adding figures for the materials and methods. Add one workflow diagram at the beginning of the materials and method part to show their project.
Results and discussion part I would recommend adding a table that compares pre-emergence herbicides for sunflower in the literature based on yield reduction, time of application, environmental conditions, and soil texture.
Author Response
Reviewer #3:
The result of the article titled “Weed control and selectivity of herbicide applied in pre-emergence with a backpack sprayer on sunflower ” is interesting.
To improve the manuscript It would be better if the title were revised. The tile does not need to include backpack sprayers, for example.
The title has been changed.
The introduction should be improved. They should compare the pre-emergence herbicide to the post-emergence herbicide for sunflower, explaining why each is better in the location. The authors need to explain why these S-metolachlor, sulfentrazone, flumioxazin, and trifluralin herbicides were chosen and explain what is the initial phase.
The introduction has been corrected.
Materials and methods: The author should explain when herbicide was applied (how many days after or before planting), the method of injury evaluation (crop? weed? visual), and yield harvesting. I would recommend adding figures for the materials and methods. Add one workflow diagram at the beginning of the materials and method part to show their project.
The material and methods were corrected, but we did not agree to insert a workflow diagram, as the descriptions are in the text.
Results and discussion part I would recommend adding a table that compares pre-emergence herbicides for sunflower in the literature based on yield reduction, time of application, environmental conditions, and soil texture.
The descriptions of the studies used in the discussions are in the text, so we chose not to insert a table, since the study is not of critical weed control period.
Round 2
Reviewer 1 Report
Although much effort has been put in the introduction, little has been done to address my issues regarding how the data was analysed and represtend. To illustrate this please see Table 2 as example - attached PDF
Note the “d” and “c” allocations to the control without weeding treatemts– all have values of “0”. These values can only differ significantly if each cultivar was analysed separately per clay soil. If this was the case, then this must indicated as such in the M&M, so that the reader can then understand that because this was done in this matter no comparisons between the two cultivars can then be made within the respective clay soils, nor can comparisons be made directly between the clay and sandy soils.
As my main concerns have not been satisfactory addressed, I can again not agree that the study is scientifically sound enough to published, as correct representation of data is the foundation of scientific publication. My suggestion would be that the article be reviewed first by a biostatatisian, after which it can be sent back to weed scientists for review. Clarity will accordingly be provided as to the best way to analyse and present the data, in case I am fully mistaken regarding what was done in this article.
Comments for author File: Comments.pdf
Author Response
Dear Reviewer#1,
The experimental design used was a randomized block design with a 2 x 6 factorial scheme with four repetitions, analyzing two cultivars (M 734 and CF 101) and six control methods (four chemicals and two controls, one without application and the other with weeding), in two soils with different textures (clayey and sandy), with independent experiments in each soil type. See in item M&M.
Since there was no interaction between the factors, the mean test is only for the treatments in the column. See the correction in the footnote of the tables.
In the text, the soil types were not compared, as they are independent studies.
The text is well explained in the Results item, see below:
There was no interaction (F = 0.55, p > 0.05) among the weed control for factors (cultivars and treatments) at 21 and 28 DAA.
Regarding agronomic parameters, with respect to capitulum insertion height, there was no interaction (F = 0.41, p > 0.05) among the factors (cultivars and treatments).
Regarding to stem diameter, there was no interaction (F = 0.92, p > 0.05) among the factors (cultivars and treatments).
For capitulum diameter, there was no interaction (F = 0.80, p > 0.05) among factors (cultivars and treatments).
Regarding the weight of 1,000 achenes, there was no interaction (F = 0.12, p > 0.05) among the factors (cultivars and treatments).
In sandy soil, there was a difference only in relation to the cultivar M 734 in plants submitted to treatment with S-metolachlor, which stood out with the highest average (1,291 kg ha-1), different from plants from the control without weeding, which had the lowest average yield (1,246 kg ha-1) with no interaction between the factors (cultivars and treatments) (F = 0.98, p > 0.05).
Best regards,
