Establishment of a Callus-Based Regeneration System for Lilium regale

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript presents a clear and generally well-designed callus-based regeneration system for Lilium regale, with appropriate experimental structure and thorough documentation of media compositions, culture conditions, and quantitative outcomes.

Major comments

1. Clarify novelty and positioning: The introduction convincingly motivates regale as a valuable species but does not clearly distinguish how this regeneration system advances beyond prior lily callus systems, especially given existing work on L. regale and other lilies.

Please try to explain (a) what was not previously available for L. regale (e.g., no callus-based system vs. only bulblet-based methods; no combined optimization of induction, expansion, and redifferentiation, etc.). And (b) how the induction-contamination line approach provides a more generalizable or rigorous optimization strategy than conventional “highest induction rate” selection.

2. In the propagation section, the statement that the 6-BA/NAA ratio in PM5 “could activate the expression of cell cycle-related genes more efficiently” is not experimentally supported in this study. Similar mechanistic language about PIC toxicity and gene-level explanations goes beyond the presented data, which are purely phenotypic (rates, morphology). It was recommended to the author that they rephrase such sentences to descriptive terms and cite relevant literature rather than implying direct evidence from this work. Avoid using strong mechanistic verbs like “activate the expression” unless gene expression assays are provided.
3. Scope and strength of conclusions: The data convincingly demonstrate a robust regeneration system for L. regale from scales, through callus, to acclimatized plants. The final sentence about “transgenetic studies of lilies” should explicitly be framed as a potential or future application.

Minor comments

1. Ensure consistent use of terms: sometimes “bulbs,” sometimes “bulbils,” “small bulbs,” “balls,” and “callus spheres” appear; standardizing terminology will reduce confusion.
2. The acclimatization protocol is well detailed; consider specifying the environmental conditions during acclimatization (temperature, relative humidity, and light) if possible.
3. The abbreviation list is helpful; make sure every abbreviation (e.g., DR, CPR) is defined at first mention in the main text as well as in the list.

Author Response

Comments 1.0 
Reviewer #1: This manuscript presents a clear and generally well-designed callus-based regeneration system for Lilium regale, with appropriate experimental structure and thorough documentation of media compositions, culture conditions, and quantitative outcomes.

Responses 1.0

Thank you very much for your suggestions. We appreciate your recognition of our research.

 

Comments 1.1

Clarify novelty and positioning: The introduction convincingly motivates regale as a valuable species but does not clearly distinguish how this regeneration system advances beyond prior lily callus systems, especially given existing work on L. regale and other lilies.

Responses 1.1

Thank you very much for your suggestions. L. regale represents the assembled genome within the genus Lilium, and the establishment of a callus system holds significant importance for subsequent genetic transformation studies in lilies. We added the following to lines 80-85 of the paper to indicate that our research is of pioneering significance.

The first genome assembly of Lilium regale was released recently [4]. This achievement significantly filled a gap in research concerning the assembly and functional analysis of mega-genomes in plants, while also provided crucial references for lily genetic evolution and molecular breeding. However, a complete regeneration system based on callus tissue has yet to be established, which limits the accelerated advancement of lily variety improvement.

 

Comments 1.2

Please try to explain (a) what was not previously available for L. regale (e.g., no callus-based system vs. only bulblet-based methods; no combined optimization of induction, expansion, and redifferentiation, etc.).

Responses 1.2

Thank you very much for your suggestions. In Comments 1.1, we discussed the innovative nature of our establishment of the L. regale callus and regeneration system. This study also represents the first optimization of the induction, expansion, and redifferentiation media for L. regale scale callus. However, the lily genome is exceptionally large and exhibits strong gene dependency. Consequently, we undertook numerous attempts to preliminarily evolve the callus regeneration system, successfully inducing L. regale callus. Moreover, in the Discussion on the lines 280-290, we have also addressed this issue. Below is the original text.

Asexual reproduction is currently a highly efficient and rapid method for propagating ornamental plants, significantly shortening the plant’s development cycle [29]. In particular, the propagation method involving the formation of adventitious buds through the redifferentiation of callus tissue has become the mainstream approach for asexual reproduction in ornamental plants [30]. Currently, most research on the asexual propagation of Lilium regale focuses on directly inducing adventitious buds from bud tissues or scales to develop into new plants, with limited studies exploring the route of plant formation through the redifferentiation of callus tissues [24,31]. This experiment used Lilium regale scales as explants to induce callus tissue through thin-layer culture. Subsequent callus expansion and redifferentiation yielded a large number of adventitious buds and establishing an effective regeneration system for efficient propagation of Lilium regale.

 

Comments 1.3

And (b) how the induction-contamination line approach provides a more generalizable or rigorous optimization strategy than conventional “highest induction rate” selection.

Responses 1.3

Thank you very much for your suggestions. During lily production, factors such as continuous cropping obstacles often result in bulbs harbouring viruses and fungi, leading to culture medium contamination during in vitro propagation. Consequently, the optimal callus induction protocol in current experiments may be accompanied by higher contamination levels. We therefore introduce the “induction-contamination equilibrium line” as a multi-indicator screening strategy, providing a novel approach for establishing efficient, low-contamination regeneration systems. Additionally, due to the limited environmental conditions of this trial, we have added a section on lines 373-378 of the Discussion section to explain that IC lines may introduce error influences. The revised content is as follows.

Due to the limited number of treatments in this screening experiment, no statistically significant relationship between induction rate and contamination rate could be established. The application of the IC line method here primarily supports intuitive comparison and comprehensive decision-making. Future research may employ more comprehensive experimental designs with increased treatment numbers to analyze the complex relationships between factors and contamination rates in greater depth.

 

 

Comments 1.4

In the propagation section, the statement that the 6-BA/NAA ratio in PM5 “could activate the expression of cell cycle-related genes more efficiently” is not experimentally supported in this study.

Responses 1.4

Thank you very much for your suggestions. We consider your question to be of great importance to our research and sincerely apologize for the inadequate description in this section. In the new revised version, we have removed this section and added a discussion on lines 325-332 regarding the impact of plant growth regulator (6-BA and NAA) configuration ratios on callus expansion. The following is our supplementary content.

  Callus tissue can redifferentiate under specific conditions to form new plant organs, thereby developing into complete plant individuals [43]. Current research indicates that the direction of callus redifferentiation is controlled by the concentration ratios of different plant growth regulators [44]. Normasari et al. [45] found that at a 6-BA concentration of 0.5 mg/L, the callus tissue of Pogostemon cablin exhibited increased biomass growth as NAA content rose. This experiment found that the plant growth regulator ratio of 6-BA (1.00 mg/L) to NAA (0.5 mg/L) in PM5 medium effectively promotes callus expansion.

 

 

 

Comments 1.5

Similar mechanistic language about PIC toxicity and gene-level explanations goes beyond the presented data, which are purely phenotypic (rates, morphology). It was recommended to the author that they rephrase such sentences to descriptive terms and cite relevant literature rather than implying direct evidence from this work. Avoid using strong mechanistic verbs like “activate the expression” unless gene expression assays are provided.

Responses 1.5

Thank you very much for your suggestions. We have amended the overly speculative content in the article. We have amended lines 319–324 of the original text to state that the addition of PIC markedly reduced the ability of scales to redifferentiate (Figure 2c). The following are the revised contents.

PIC was also identified as an effective herbicide, primarily functioning by disrupting plant cells and inhibiting plant growth [39,40]. As shown in Figure 2c, in IM2, IM3, IM4, IM5, IM7 and IM9 with addition of PIC, scales showed reduced redifferentiation capabilities compared to in IM1, IM6, and IM8 without PIC, suggesting that PIC helped to maintain the callus status of scales in Lilium regale. These results were consistent with previous reports of PIC inhibiting organ differentiation in other plant species [41,42].

 

 

Comments 1.6

Scope and strength of conclusions: The data convincingly demonstrate a robust regeneration system for L. regale from scales, through callus, to acclimatized plants. The final sentence about “transgenetic studies of lilies” should explicitly be framed as a potential or future application.

Responses 1.6

Thank you very much for your suggestions. We have amended the content of lines 393–396 of the original text, the following are the revised contents. 

Our results also supported that new seedlings could be reproduced by our callus-based regeneration system. These offered new insights into propagation of Lilium regale and provided potential materials for transgenic studies of lilies in the future.

 

Comments 1.7

Ensure consistent use of terms: sometimes “bulbs,” sometimes “bulbils,” “small bulbs,” “balls,” and “callus spheres” appear; standardizing terminology will reduce confusion.

Responses 1.7

Thank you very much for your suggestions. We have re-edited the entire text and standardized the terminology throughout , unified all references to “bulbs”.

 

 

Comments 1.8

The acclimatization protocol is well detailed; consider specifying the environmental conditions during acclimatization (temperature, relative humidity, and light) if possible.

Responses 1.8

Thank you very much for your suggestions. Variations may occur during transplantation due to uncontrollable environmental factors, but we tried to ensure optimal conditions throughout the process. We added environment parameters on lines 196-198, the following are the revised contents.

All tests were conducted under identical environmental conditions (i.e., temperature: 25±2°C, humidity: 70–80%, and photoperiod: 12 hours light/12 hours dark).

 

 

 

Comments 1.9

The abbreviation list is helpful; make sure every abbreviation (e.g., DR, CPR) is defined at first mention in the main text as well as in the list.

Responses 1.9

Thank you very much for your suggestions. We have re-edited the entire text, supplementing the full form of abbreviations upon their first appearance in the main body. To ensure consistency throughout the article, we have converted all such abbreviations to italics.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors provide an analysis of the effects of plant growth regulator combinations on Lilium propagation via scales. The introduction has an appropriate background and rational for the technique, which is widely used for Lilium cultivars. The authors experimental design and subsequent analysis, however, does not support the designation of an optimal PGR composition for the system investigated.

In particular, the design associated with callus induction uses 9 different media as factors and performs a one-way ANOVA for responses that include induction rate, contamination rate and (re)differentiation rate. Since the media are treated as categorical, there are no statistical parameters that can be extracted to identify the specific roles of 6BA, NAA or PIC and any designation of optimal PGR combinations is simply an outcome that aligns with the the highest or lowest value for the experimental response. The statistics, as presented by the authors, simply state that certain culture combinations are different. It is then trivial to designate the highest or lowest value as optimal. The post-ANOVA analysis also indicate that culture IM5 (designated 'optimal' by the authors) is not statistically different than cultures IM6, IM8 or IM9 (they all have superscript 'a' or 'b').

Some points to consider:

1. The orthogonal design shown in Tables 1 and 2 is not explicitly used in any statistical analysis since the factors 6BA, NAA and PIC are NOT analyzed as 3 different effects. It is therefore misleading to the reader to designate 'orthogonal' as a special condition used in this work. Indeed, given 3 replicates of 9 runs, if the data for CIR, for example, were analyzed with PGRs as quantitative factors using a main effects plus interaction model, there are no statistically distinguishable terms. The authors need not state the design proposed is orthogonal, except perhaps in passing. Tables 1 and 2 are redundant and need not both be used.
2. Table 3 is missing the column for NAA.
3. Table 5 seems to plot the same set of data 3 different times and, not surprisingly, arrives at the same outcome each time. There are only 9 different media combinations and the statistical model will predict a value for CIR or CCR for each media shown. The introduction of a relationship between responses is not rigorous since the statistics for thea straight line fit are not provided. The statistics are easily found, however, and indicate that the slope is not different from '0'. This means contamination does not change with increasing induction rates, contrary to the authors use of 'IR" as an 'optimal' indicator.
4. The abbreviation PIC is not identified in the bulk of the manuscript until line 310 after it was used several times.
5. All references to 'optimal' should be removed from the text. 

Additional comments. The authors have used 27 runs in their design with 3 levels of 3 factors. The initial impression in skimming the manuscript was that this would allow for quantitative analysis using a response surface model for three factors. In such a case, an optimal outcome could be anticipated. Their design (a 1/3 3³ fractional factorial), however, does not support quadratic terms. Better designs are found in the literature (CCDs or Box-Behnken) that result in fewer experiments (say 15) to achieve a rigorous response surface model and ideas on combining responses using 'desirability' functions can work to choose a single PGR combination. I mention this only because the authors hinted at the significance of an orthogonal design so they are aware that some designs are better than others.

 

Author Response

Comments 2.0
Reviewer #2: The authors provide an analysis of the effects of plant growth regulator combinations on Lilium propagation via scales. The introduction has an appropriate background and rational for the technique, which is widely used for Lilium cultivars. The authors experimental design and subsequent analysis, however, does not support the designation of an optimal PGR composition for the system investigated.

Responses 2.0

Thank you very much for your insightful and professional comments on the statistical methods employed in this paper.

 

Comments 2.1

In particular, the design associated with callus induction uses 9 different media as factors and performs a one-way ANOVA for responses that include induction rate, contamination rate and (re)differentiation rate. Since the media are treated as categorical, there are no statistical parameters that can be extracted to identify the specific roles of 6BA, NAA or PIC and any designation of optimal PGR combinations is simply an outcome that aligns with the the highest or lowest value for the experimental response. The statistics, as presented by the authors, simply state that certain culture combinations are different. It is then trivial to designate the highest or lowest value as optimal.

Responses 2.1

Thank you very much for your suggestions，We fully understand and concur with your observations. We acknowledge that within the current statistical framework, single-factor analysis of variance (ANOVA) combined with Duncan's test can only determine whether significant differences exist between different medium treatment groups. It cannot quantify the independent contribution of each hormone factor (6-BA, NAA, PIC) and their interactions to the response variables (CIR, CCR, DR). Directly labelling the medium with the highest induction rate (e.g., IM9) or the medium selected via the IC line (IM5) as ‘optimal’ is indeed imprecise in the strict sense of statistical modelling. This is because such labelling is based on direct comparisons of observed values rather than theoretical optimal points derived through response surface modelling (e.g., Response Surface Methodology, RSM).

The primary purpose of employing a three-factor, three-level orthogonal design (L₉(3⁴)) was to conduct an efficient preliminary screening within a limited experimental scale (nine treatment combinations). This aimed to explore the potentially suitable concentration ranges of three key hormones for callus induction in Lilium regale, rather than establishing precise dose-response mathematical models.

As you astutely observed, such orthogonal designs (1/3³ factorial analysis designs) typically lack sufficient experimental data points to reliably estimate quadratic terms (curvature) and all interaction terms. Consequently, they are not entirely suitable for constructing comprehensive response surface models (such as Central Composite Designs (CCDs) or Box-Behnken designs). Our initial research objective was to identify one or several practical culture medium formulations achieving a favourable balance between induction rate, contamination rate, and redifferentiation rate, thereby providing a starting point for establishing a comprehensive regeneration system. Hence, we selected this highly efficient screening design.

In response to your feedback, we shall incorporate the following substantive amendments into the revised draft.

We replaced the term ‘optimal’ by replacing ‘optimal medium’ throughout the text with ‘a suitable medium under the experimental conditions’ and to clearly state that this study is designed as a ‘preliminary screening experiment’. Supplement the explanation by noting that, due to limitations in the number of experimental points, no factorial effect quantification analysis (such as regression coefficient estimation) was conducted for 6-BA, NAA, and PIC, acknowledging this as a limitation of the study.

 

 

 

 

Comments 2.2

The post-ANOVA analysis also indicate that culture IM5 (designated 'optimal' by the authors) is not statistically different than cultures IM6, IM8 or IM9 (they all have superscript 'a' or 'b').

Responses 2.2

Thank you very much for your suggestions，your observation is entirely correct, and this constitutes a significant correction to the data interpretation. According to our comparative results, there is indeed no statistically significant difference in induction rates between culture media IM5 (induction rate 63.3%), IM6, IM8, and IM9 (induction rate 66%) (all belonging to the same letter groups “a” and “b”). To designate IM5 as “optimal” based solely on the “highest induction rate” metric would be statistically unreliable, as IM9 exhibits numerically higher induction rates without statistical significance. Nevertheless, our recommendation of IM5 as the medium with the best overall performance is grounded in a more comprehensive decision-making framework, rather than a single-factor comparison of induction rates alone. In plant tissue culture, establishing an efficient regeneration system demands not only high induction rates but also prioritizes low contamination rates as an equally critical objective. High contamination rates directly lead to experimental failure, resource wastage, and irreproducible results. This is precisely why we introduced the IC line method. This approach aims to simultaneously optimize both induction rate and contamination rate – two objectives that often present a trade-off.

We shall make the following amendment to the full text, replacing all instances of the description ‘optimal’ with ‘a suitable medium under the experimental conditions’.

 

 

 

Comments 2.3

The orthogonal design shown in Tables 1 and 2 is not explicitly used in any statistical analysis since the factors 6BA, NAA and PIC are NOT analyzed as 3 different effects. It is therefore misleading to the reader to designate 'orthogonal' as a special condition used in this work. Indeed, given 3 replicates of 9 runs, if the data for CIR, for example, were analyzed with PGRs as quantitative factors using main effects plus interaction model, there are no statistically distinguishable terms. The authors need not state the design proposed is orthogonal, except perhaps in passing. Tables 1 and 2 are redundant and need not both be used.

Responses 2.3

Thank you very much for your suggestions，we have provided a detailed explanation of this issue in comment 2.1. We accept your suggestion and will tone down the emphasis on the specificity of ‘orthogonal design’ in the revised manuscript, describing it merely as a ‘systematic screening design’ (on line 121、 291 and 358). We will also explicitly state that it is primarily used to narrow the scope of optimization.

 

 

 

Comments 2.4

Table 3 is missing the column for NAA.

Responses 2.4

Thank you very much. We have carefully considered the issue you raised, but we have been unable to identify the root cause. Table 3 indeed include the column of NAA (see below).

 

 

 

 

 

Comments 2.5

Table 5 seems to plot the same set of data 3 different times and, not surprisingly, arrives at the same outcome each time. There are only 9 different media combinations and the statistical model will predict a value for CIR or CCR for each media shown. The introduction of a relationship between responses is not rigorous since the statistics for thea straight line fit are not provided. The statistics are easily found, however, and indicate that the slope is not different from '0'. This means contamination does not change with increasing induction rates, contrary to the authors use of 'IR" as an 'optimal' indicator.

Responses 2.5

Thank you very much for your suggestions，we recognize that this is a very serious issue. To ensure data accuracy, we have included the data characteristics of the fitted IC line (Y = -0.1299 × X + 24.88) in the annotations of Figure 5.

The scatter plots in Figures 5a, 5b, and 5c indeed represent the same nine sets of paired data (CIR, CCR), with colour and shape grouping based solely on the concentration of one hormone. Our original intention was to visually demonstrate the distribution patterns of data points under varying hormone concentrations through this approach. However, this presentation has a risk of conveying the misconception of ‘duplicate plotting’. We shall consolidate and optimize this representation.

We fully concur that the IC line should not be interpreted as a statistically significant model with predictive capabilities. Nevertheless, we maintain that it retains value within this study as a heuristic, visualized multi-criteria decision-making support tool.

We revised relevant descriptions in Discussion and added a section on lines 371-378 to Discussion, as below.

Based on a dual-indicator comprehensive assessment of induction rate and contamination rate, IM5 demonstrated the best equilibrium among the nine media tested in this trial. Due to the limited number of treatments in this screening experiment, no statistically significant relationship between induction rate and contamination rate could be established. The application of the IC line method here primarily supports intuitive comparison and comprehensive decision-making. Future research may employ more comprehensive experimental designs with increased treatment numbers to analyse the complex relationships between factors and contamination rates in greater depth.

 

 

 

Comments 2.6

The abbreviation PIC is not identified in the bulk of the manuscript until line 310 after it was used several times.

Responses 2.6

Thank you very much for your suggestions，We shall proofread the manuscript once more to ensure that proper nouns are given their full names upon their first appearance. In this revision, we have added the full names of PICs and standardized the initial description format for other plant growth regulators (TDZ, NAA, 6-BA). The following are the revised contents (on lines 73-80).

Current research indicates that Picloram (PIC) is a key factor essential for the induction and maintenance of embryonic callus tissue in lily (Lilium) species during tissue culture [24,25]. Using PIC to induce callus formation in 33 lily genotypes, 30 genotypes were capable of forming callus tissue [17]. Using combinations of Thidiazuron (TDZ), 1-Naphthaleneacetic acid (NAA), or 2,4-Dichlorophenoxyacetic acid (2,4-D) and 6-Benzylaminopurine (6-BA) only induce the formation of bulbs in Lilium longiflorum, whereas 2 mg/L PIC effectively promotes the development of somatic embryos [26].

 

 

Comments 2.7

All references to 'optimal' should be removed from the text.

Responses 2.7

Thank you very much for your suggestions，we shall proofread the paper once more and amend similar phrasing. As mentioned in response 2.2, we have replaced all instances of “optimal” with “a suitable medium under the experimental conditions”.

 

 

Comments 2.8

Additional comments. The authors have used 27 runs in their design with 3 levels of 3 factors. The initial impression in skimming the manuscript was that this would allow for quantitative analysis using a response surface model for three factors. In such a case, an optimal outcome could be anticipated. Their design (a 1/3 33 fractional factorial), however, does not support quadratic terms. Better designs are found in the literature (CCDs or Box-Behnken) that result in fewer experiments (say 15) to achieve a rigorous response surface model and ideas on combining responses using 'desirability' functions can work to choose a single PGR combination. I mention this only because the authors hinted at the significance of an orthogonal design so they are aware that some designs are better than others.

Responses 2.8

Thank you very much for your suggestions，you have not only accurately identified the theoretical limitations of our current experimental design (1/3 3³ factorial analysis), but also pointed us—and subsequent researchers in this field—towards a more rigorous and robust optimisation pathway. We are deeply grateful for this and fully concur with your perspective. The primary objective of this study was to establish a stable, operational callus regeneration system for Lilium regale. This constitutes foundational work, building from scratch. Consequently, our foremost requirement was to efficiently screen, within a controllable experimental scale, the range of hormone formulations capable of successfully inducing, amplifying, and redifferentiating callus tissue. Within this context, the three-factor, three-level orthogonal design (L₉(3⁴)) serves as a classical and highly efficient screening tool. It systematically explores the combined effects of three factors at three levels with minimal experimental runs (nine), making it ideally suited for our ‘preliminary exploration and screening’ research phase. This approach enabled us to rapidly identify high-performing formulations such as IM5 and PM5, successfully achieving the core objective of establishing the regeneration system.

We fully concur that our experimental design indeed has limitations if the research objective is to precisely quantify and model hormone effects while seeking theoretically optimal solutions. To appropriately address your suggestion, we shall incorporate the following adjustments in the revised manuscript: adding a section on limitations and future prospects within on the lines 356-366 of the Discussion section. The following is newly added content in the discussion section. 

This study successfully established an in vitro regeneration system for Lilium regale based on scale callus tissue. It should be noted, however, that the three-factor, three-level systematic screening design employed primarily served the preliminary objective of rapid screening. This design inherently has limitations in precisely quantifying the independent effects, interactions, and non-linear responses of plant growth regulators (6-BA, NAA, PIC). Future research may build upon the effective concentration ranges identified herein by employing more sophisticated experimental designs (such as central composite or Box-Behnken designs), combined with response surface analysis and multi-objective fitness function optimisation methods. This approach would enable the construction of more predictive mathematical models and elucidate the precise modes of action for each factor during callus induction, proliferation, and redifferentiation processes.

 

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

the authors have provided revisions that align with the previous comments from this reviewer. I would recommend the following adjustments to the manuscript that avoid redundant information

remove Table 1 and the parenthetical reference to Table 1 (line 121) since the information is easily found in Table 2.

remove the figure legends in Figures 1-4 (right hand side array of colored symbols and notation)  since it is repeated on the x-axis. This makes the figure easier to read.

Author Response

Comments 2.0
Reviewer #2: the authors have provided revisions that align with the previous comments from this reviewer. I would recommend the following adjustments to the manuscript that avoid redundant information

Responses 2.0

Thank you very much for your suggestions. We have revised the inappropriate content in the paper according to the reviewers' comments.

 

 

 

Comments 2.1 

remove Table 1 and the parenthetical reference to Table 1 (line 121) since the information is easily found in Table 2.

Responses 2.1

Thank you very much. We find your suggestions highly valuable. In the new revision, we have removed Table 1 and all references to it in that section. Additionally, we have renumbered Tables 2 and 3 as Tables 1 and 2, respectively, and corrected their citations accordingly. The following outlines the modifications made to the text.

On lines 120-122: Three concentration gradients were set respectively and systematic screening tests with three factors and three levels were carried out, which formed nine different mediums as listed in Table 1.

On lines 155-156:there were 6 PGRs formulas in total (Table 2).

On lines 170-171: Firstly, the callus induction rate and contamination rate values of different mediums in Table 1 were plotted in a scatter plot.

On lines 200-201: As shown in Figure 1a, ten explants were put into different mediums as defined in Table 1.

 

 

 

 

Comments 2.2 

remove the figure legends in Figures 1-4 (right hand side array of colored symbols and notation)  since it is repeated on the x-axis. This makes the figure easier to read.

Responses 2.2

Thank you very much for your suggestions. We fully agree with your perspective. The legends in Figures 1-4 indeed present potential redundancy. Therefore, we have removed the relevant content from Figures 1-4 in the new revised version. Below is the revised version of Figures 1-4.

 

 

 

 

Fig1                                fig2

 

Fig3                                    fig4              

 

Author Response File: Author Response.pdf