Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

A well-designed experiment and a well-written manuscript. I have just a few minor remarks:

1. End of the first paragraph of ‘Introduction’: shouldn’t the term ‘intra-specific cannibalism’ simply be replaced with the term ‘cannibalism’. By definition, cannibalism is always intra-specific and the interspecific interaction resembling cannibalism is IGP.

2. The sentence near the end of p. 2 [Optimization of the oviposition sites in systems with large generation times ratio (aphids having extremely short life cycle, compared e.g. to coccids [17] then favor using compounds deposited by conspecifics as a clue to search for an alternative food patch [18].] is not entirely clear to me and probably needs to be rewritten. What does ‘generation times ratio’ mean? Shouldn’t the ratio be high instead of large? There is no closing bracket, so it is unclear where the parenthetical remark ends.

3. Reference no. 4: change ‘Synder’ to ‘Snyder’

4. The same paper by Reznik et al. is quoted in the text and listed in ‘Reference’ twice: as [22] and [26].

5. Figures 1 and 2: Is there any justification for connecting the averages for individual densities with lines?

Author Response

A well-designed experiment and a well-written manuscript. I have just a few minor remarks:

1. End of the first paragraph of ‘Introduction’: shouldn’t the term ‘intra-specific cannibalism’ simply be replaced with the term ‘cannibalism’. By definition, cannibalism is always intra-specific and the interspecific interaction resembling cannibalism is IGP.

Response: sure, corrected

2. The sentence near the end of p. 2 [Optimization of the oviposition sites in systems with large generation times ratio (aphids having extremely short life cycle, compared e.g. to coccids [17] then favor using compounds deposited by conspecifics as a clue to search for an alternative food patch [18].] is not entirely clear to me and probably needs to be rewritten. What does ‘generation times ratio’ mean? Shouldn’t the ratio be high instead of large? There is no closing bracket, so it is unclear where the parenthetical remark ends.

Response: Devided to two sentences for clarity.

3. Reference no. 4: change ‘Synder’ to ‘Snyder’

Response: Appologies to Snyder.

4. The same paper by Reznik et al. is quoted in the text and listed in ‘Reference’ twice: as [22] and [26].

Response: References renumbered

5. Figures 1 and 2: Is there any justification for connecting the averages for individual densities with lines?

Response: it is not necessary, but density can be regarded either as categorial or continuous predictor, that can connect datapoints; Anyway, connecting lines deleted.

Reviewer 2 Report

Comments and Suggestions for Authors

The practical and evolutionary implications of larval tracks on conspecific ladybirds is important. While many aspects of this phenomenon have been investigated, the present study explores new aspects with regard to developmental time of fourth instars and pupae and adult body mass in relation to track pheromones and larval density. Thus, the results of the study could fill important knowledge gaps.

However, the analysis of the study was improper, and the results are not reliable. Here are some of the items to address, concerned principally with the analyses.

There were three larval densities used in this study. Each of the larvae within the 4-larvae and 8-larvae treatments were grouped within a Petri dish, which was the experimental unit. Therefore, individuals of the four or eight larvae within each Petri dish represented subsamples, and the value used for analysis should have been the average for each of the groups, whether data were derived for 4th instars, pupae, or adults. However, it appears that the authors incorrectly considered each individual within a Petri dish (=experimental unit) as a replicate, and this resulted in pseudoreplication and in inflated degrees of freedom, invalidating the ANOVA results. 

The ANOVA was inadequately described, in that the authors do not address whether the treatments were considered random vs. fixed, and how (or whether) the analysis handled the three replications over two months, i.e., rep considered fixed or random. Thus, aspects of the ANOVA validity could not be corroborated.

In addition, it appears that the df of the ANOVA were incorrect for the larval density treatments, i.e., n = 3 for the three larval densities (1, 4, and 8 larvae per Petri dish). The n = 3 treatments should have resulted in df = 2 = n-1. However, for developmental time and adult body mass, the authors report df = 1 for density treatments. Strangely, for pupal developmental time, the authors report only a single ANOVA result, stating that there “was no significant difference in the pupal developmental time between the six treatments (F2,348=1.40, p=0.23)...,” with the one result reported here with df = 2. This does not make sense. 

Although the ANOVA were conducted with pseudoreplicates and are invalid, comments are needed about Fig. 1 and about the reporting of body mass results, as follows. 

Fig. 1. Letters above each circle and triangle depicting the LSD post hoc results would have been appropriate and helped the reader follow the results. 

Body mass. In this instance, pheromone/clean treatment and density were significant, but their interaction was not. Thus, one reports the effect of density across pheromone/clean treatments, and not separately for C and P. Similarly, it is appropriate to compare C vs. P across densities, and it is not appropriate to compare, e.g., C1 vs. P1, and so forth.

Statistical consultation may be advisable.

I did not read the Abstract, Discussion, and Conclusions sections, as the statistical analyses were invalid, and any inferences and general principles based on them were unreliable. 

The manuscript needs to cite Kumar B, Mishra G, Omkar. Larval and female footprints as feeding deterrent cues for immature stages of two congeneric ladybird predators (Coleoptera: Coccinellidae). Bulletin of Entomological Research. 2014;104(5):652-660. doi:10.1017/S0007485314000479. This study showed that larval tracks prolonged total developmental duration of ladybirds.

Comments on the Quality of English Language

I only read Intro and M&M. Both are wordy at times, e.g. first sentence contains "globally...and also known worldwide." Second paragraph reads clumsily "The larvae go through four larval instars...."

The Introduction contained salient background points, but its structure and organization need improvement. Here are a few items to help improve the Intro: 

Break the second paragraph apart, beginning a new paragraph with the sentence beginning as “Both eggs and pupae are immobile,....” The mini-paragraph beginning with “Besides cannibalism” should be combined with the new (third) paragraph. 

In the new paragraph, “H. axyridis" needs to be defined here rather than >15 lines later. 

The paragraph beginning as “Exposure of either larvae or adults...” lacks a topic sentence. 

The last two “paragraphs” are underdeveloped conceptually, as they consist of only one and two sentences, respectively. (I realize one tells the reader what will be conducted in this study.) They should be re-constructed. 

Author Response

Reviewer 2

However, the analysis of the study was improper, and the results are not reliable. Here are some of the items to address, concerned principally with the analyses.

There were three larval densities used in this study. Each of the larvae within the 4-larvae and 8-larvae treatments were grouped within a Petri dish, which was the experimental unit. Therefore, individuals of the four or eight larvae within each Petri dish represented subsamples, and the value used for analysis should have been the average for each of the groups, whether data were derived for 4th instars, pupae, or adults. However, it appears that the authors incorrectly considered each individual within a Petri dish (Response=experimental unit) as a replicate, and this resulted in pseudoreplication and in inflated degrees of freedom, invalidating the ANOVA results. 

Response: Thanks for suggesting correct analysis. We now used the Petri dish as the experimental unit (we calculated dish means) to avoid pseudoreplication, and included the dish ID as random effect.

The ANOVA was inadequately described, in that the authors do not address whether the treatments were considered random vs. fixed, and how (or whether) the analysis handled the three replications over two months, i.e., rep considered fixed or random. Thus, aspects of the ANOVA validity could not be corroborated.

Response: Now we define the mixed model ANOVA with two fixed effects and one random effect. The three replications were rather a continuum of dishes in time, so that the dish ID describes this effect fully.

In addition, it appears that the df of the ANOVA were incorrect for the larval density treatments, i.e., n Response= 3 for the three larval densities (1, 4, and 8 larvae per Petri dish). The n Response= 3 treatments should have resulted in df Response= 2 Response= n-1. However, for developmental time and adult body mass, the authors report df Response= 1 for density treatments.

Response: We appologize, it was copy/paste error. DfResponse=2.

 Strangely, for pupal developmental time, the authors report only a single ANOVA result, stating that there “was no significant difference in the pupal developmental time between the six treatments (F2,348Response=1.40, pResponse=0.23)...,” with the one result reported here with df Response= 2. This does not make sense. 

Response: Appologies for deleting other non-significant results, now completed.

Although the ANOVA were conducted with pseudoreplicates and are invalid, comments are needed about Fig. 1 and about the reporting of body mass results, as follows. 

Fig. 1. Letters above each circle and triangle depicting the LSD post hoc results would have been appropriate and helped the reader follow the results. 

Response: Letters above circles and squares added. Fortunately there are not so many significant differences.

Body mass. In this instance, pheromone/clean treatment and density were significant, but their interaction was not. Thus, one reports the effect of density across pheromone/clean treatments, and not separately for C and P. Similarly, it is appropriate to compare C vs. P across densities, and it is not appropriate to compare, e.g., C1 vs. P1, and so forth.

Response: Comparisons C1xP1 and C8xP8 omitted.

I did not read the Abstract, Discussion, and Conclusions sections, as the statistical analyses were invalid, and any inferences and general principles based on them were unreliable. 

Response: We made the required mixed model analysis and that revealed the same fixed effects as they were in the original version of the manuscript. Results and discussion did not chnaged.

The manuscript needs to cite Kumar B, Mishra G, Omkar. Larval and female footprints as feeding deterrent cues for immature stages of two congeneric ladybird predators (Coleoptera: Coccinellidae). Bulletin of Entomological Research. 2014;104(5):652-660. doi:10.1017/S0007485314000479. This study showed that larval tracks prolonged total developmental duration of ladybirds.

Response: good suggestion, we added the citation and discussed why it differs from our results.

Comments on the Quality of English Language

I only read Intro and M&M. Both are wordy at times, e.g. first sentence contains "globally...and also known worldwide." Second paragraph reads clumsily "The larvae go through four larval instars...."

Response: Some corrections were made throughout.

The Introduction contained salient background points, but its structure and organization need improvement. Here are a few items to help improve the Intro: 

Response: we made the proposed changes, rearrangements, additions…

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript investigates whether larval footprint chemicals in Harmonia axyridis can mitigate the development delay and reduced adult body mass caused by high larval density. Using a factorial design with three densities (1, 4, 8 larvae per dish) and two chemical environments (daily dish replacement vs. footprint accumulation), the study addresses ecologically and chemically relevant questions. The experimental design is simple and the sample size is relatively large. However, the manuscript currently has substantial issues in mechanistic interpretation, statistical rigor, independence of experimental units, data transparency, and figure/table presentation. Major revision is required before the manuscript can be considered for publication. Detailed comments follow.

1. Overstated inference regarding “development-enhancing pheromone”. The authors refer to the footprint chemicals as a “development-enhancing pheromone.” However, the results only show that the P group did not exhibit developmental delay under high-density conditions. This does not demonstrate developmental acceleration, nor does it distinguish whether the effect is “enhancing” or merely buffering density-induced stress. Please adjust wording to “a potential buffering effect that may mitigate density-induced developmental delay,” and explicitly state in the Discussion that this interpretation is speculative, as no feeding, metabolic, or hormone-related measurements were conducted.
2. Chemical cues were not quantified; causal claims are weak. In the P treatment, footprint accumulation is confounded with food residues, feces, and possible microbial films. Without GC-MS or behavioral assays, the origin and identity of the chemical signals remain unclear. Add a Limitation section clarifying that chemical extraction or GC-MS verification was not performed. Cite previous work quantifying Harmonia cuticular hydrocarbons to indicate future directions.
3. Lack of independence among experimental units; statistical model needs improvement. It is unclear whether larvae within the dataset originate from the same or different dishes, and no Dish_ID is provided. Treating individual larvae as independent replicates risks pseudoreplication. Provide Dish_ID in the revised dataset, and consider either performing analyses using dish means, or using mixed-effects models with Experiment or Dish_ID as a random intercept.
4. Interpretation of density effects goes beyond the support of the data.  The manuscript repeatedly suggests that high density “disturbs feeding and reduces body mass,” yet no direct measurements of feeding rate, aggression, interference behavior, or stress-related neurohormones (e.g., octopamine, AKH) were taken. Modify such statements to “possible mechanisms” and support them with relevant literature, while avoiding definitive causal claims.
5. Adult body-mass data are not available in the supplementary materials. Figure 2 presents body-mass comparisons, yet the supplementary tables do not include body-mass values. This prevents independent verification of the results. Add adult body mass (mg) to the supplementary dataset or provide it as a separate file, and specify balance precision.
6. Figure presentation issues. Figures 1 and 2 use 95% CIs while multiple comparisons use LSD, which may confuse readers. No significance letters are shown, and the color scheme is not friendly to grayscale printing. Use mean ± SE consistently and annotate significant differences with letters. Provide vector formats (EPS/TIFF) suitable for publication.
7. Additional issues that require attention: The Introduction should better articulate why the fourth instar is an important and understudied developmental stage. The 12-hour observation interval introduces interval-censoring error; if survival analysis (K-M/Cox) is not adopted, please discuss the potential impact on significance. When combining the pre-test at 23.1–23.5 °C with the main experiment at 25 °C, Experiment should be treated as a random factor. Reference items 22 and 26 are duplicated; journal abbreviations should follow ISI standards.
8. Supplementary materials need substantial improvement. The supplementary materials appear disorganized. Each table should be placed on a separate sheet, and each figure and table must include detailed captions and annotations. External OneDrive links should not be used.

This study presents an interesting ecological question and has the potential to contribute to the understanding of chemical cues in larval development. However, major issues in experimental independence, statistical analysis, data transparency, and figure/table presentation must be resolved before the manuscript can meet the publication standards of Insects.

Author Response

This manuscript investigates whether larval footprint chemicals in Harmonia axyridis can mitigate the development delay and reduced adult body mass caused by high larval density. Using a factorial design with three densities (1, 4, 8 larvae per dish) and two chemical environments (daily dish replacement vs. footprint accumulation), the study addresses ecologically and chemically relevant questions. The experimental design is simple and the sample size is relatively large. However, the manuscript currently has substantial issues in mechanistic interpretation, statistical rigor, independence of experimental units, data transparency, and figure/table presentation. Major revision is required before the manuscript can be considered for publication. Detailed comments follow.

1. Overstated inference regarding “development-enhancing pheromone”. The authors refer to the footprint chemicals as a “development-enhancing pheromone.” However, the results only show that the P group did not exhibit developmental delay under high-density conditions. This does not demonstrate developmental acceleration, nor does it distinguish whether the effect is “enhancing” or merely buffering density-induced stress. Please adjust wording to “a potential buffering effect that may mitigate density-induced developmental delay,” and explicitly state in the Discussion that this interpretation is speculative, as no feeding, metabolic, or hormone-related measurements were conducted.

Response: well formulated, we used this „buffering“ terminology.

1. Chemical cues were not quantified; causal claims are weak. In the P treatment, footprint accumulation is confounded with food residues, feces, and possible microbial films. Without GC-MS or behavioral assays, the origin and identity of the chemical signals remain unclear. Add a Limitation section clarifying that chemical extraction or GC-MS verification was not performed. Cite previous work quantifying Harmonia cuticular hydrocarbons to indicate future directions.

Response: We continue with these experiments and we have GC MS composition of tracks without contamination by feces and food. This will be another article, since it is thesis of other student.

1. Lack of independence among experimental units; statistical model needs improvement. It is unclear whether larvae within the dataset originate from the same or different dishes, and no Dish_ID is provided. Treating individual larvae as independent replicates risks pseudoreplication. Provide Dish_ID in the revised dataset, and consider either performing analyses using dish means, or using mixed-effects models with Experiment or Dish_ID as a random intercept.

Response: As also required by the reviewer 2, we made a new mixed-models ANOVA with random effects of the dishes. Fortunately, the results are basicaly the same.

1. Interpretation of density effects goes beyond the support of the data.  The manuscript repeatedly suggests that high density “disturbs feeding and reduces body mass,” yet no direct measurements of feeding rate, aggression, interference behavior, or stress-related neurohormones (e.g., octopamine, AKH) were taken. Modify such statements to “possible mechanisms” and support them with relevant literature, while avoiding definitive causal claims.

Response: some interpretations were diminished, but we also cite literature that does not support [22] but rather contradict [23] our results.

1. Adult body-mass data are not available in the supplementary materials. Figure 2 presents body-mass comparisons, yet the supplementary tables do not include body-mass values. This prevents independent verification of the results. Add adult body mass (mg) to the supplementary dataset or provide it as a separate file, and specify balance precision.

Response: There is available file: https://www.researchgate.net/publication/397940646_Body_weight_as_a_function_of_Population_Density

Response: There is already balance precision stated.

1. Figure presentation issues. Figures 1 and 2 use 95% CIs while multiple comparisons use LSD, which may confuse readers. No significance letters are shown, and the color scheme is not friendly to grayscale printing. Use mean ± SE consistently and annotate significant differences with letters. Provide vector formats (EPS/TIFF) suitable for publication.

Response: Letters are now included. Separate TIFF files were certainly submitted.

1. Additional issues that require attention: The Introduction should better articulate why the fourth instar is an important and understudied developmental stage. The 12-hour observation interval introduces interval-censoring error; if survival analysis (K-M/Cox) is not adopted, please discuss the potential impact on significance. When combining the pre-test at 23.1–23.5 °C with the main experiment at 25 °C, Experiment should be treated as a random factor. Reference items 22 and 26 are duplicated; journal abbreviations should follow ISI standards.

Response: Our 12 h interval is above usual standard daily observation in similar studies.

Response: References were renumbered.

1. Supplementary materials need substantial improvement. The supplementary materials appear disorganized. Each table should be placed on a separate sheet, and each figure and table must include detailed captions and annotations. External OneDrive links should not be used.

Response: We do not want to publish official suplementary materials within the Journal. We provide our true data sheets in Research Gate, not in OneDrive.

Response: Thanks for all constructive suggestions!

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

This is an interesting lab study. I appreciate the authors’ revision of the manuscript particularly with regard to statistical analysis. However, while notable improvement is evident in the latest version, some further revision is needed with regard to the title, Introduction, clarity of statistical methods, application of statistical analyses, and the reporting of results. I start with issues related to statistics, as these are critical.

 

Language used by the authors throughout the Methods and Results sections indicates that they have misconceptions about their own experimental set up and thus about how proceed in analyzing and reporting the results.

The authors repeatedly misspeak about treatments and treatment levels, and reference them in erroneous phrases such as “among the six treatments” or “between the six treatments” and mixing these with statements like the “treatment (C or P).” There were only two treatment factors (pheromone and larval density)—one had two levels (Clean and Pheromone) and the other had three (1, 4 and 8 larvae) –for a total of five treatment levels. However, as this was a factorial design, there were six (2 x 3) treatment combinations.

The interpretation of the significance or lack of significance of each treatment and their interaction was not consistently handled appropriately by the authors. For the effect on 4^th instar development duration, pheromone, larval density, and their interaction were all significant. Thus, it is appropriate to compare means among the individual treatment combinations using Least Square Means or something equivalent. However, this is tricky, given that the random factor of dish was highly significant. Regardless, it appears that the authors erroneously performed the Tukey HSD for each main factor but also used these within-factor Tukey results to compare means across both treatment factors, although it was not clear from the description of stat procedures in the Methods section. Thus, Fig. 1 appears incorrect.

For the effect on adult body mass, pheromone and larval density treatments were each significant; Their interaction was non-significant. Thus, differences revealed by a Tukey HSD test should have been confined to P vs. C and then separately among the 3 larval densities—all six treatment combinations should NOT have been compared against one another (e.g., “Body mass in treatment C8 was lower than in C1” is inappropriate). Thus, Fig. 2 is invalid.

The highly significant random effect of dish needs some kind of explanation, although the authors do wisely note their caution in interpreting their results in light of the random effect. Could the random effect have arisen by the way the 4^th instars were obtained for placement into the Petri dishes? That is, were faster developing larvae the first to go into the dishes? Based on the the attached Supplementary data file, could the dish effect have arisen because some of the dishes were maintained at room temperature whereas other were kept at 25 C? Experimental conditions need to be stated in the Methods.

The authors were incomplete in reporting their ANOVA results, supplying only the numerator degree of freedom and omitting the denominator d.f.

What software was (and perhaps programs were) used to conduct the statistical analyses?

 

The title is inexact, mentioning only “larval development,” whereas measurements were also made on pupae and adults in the study.

The Introduction needs improvement. Studies cited in the Introduction are referred to inconsistently in either the past or the present tense.

In the first sentence of the last paragraph of the Introduction, “investigate” needs to be corrected. The last sentence of this paragraph needs to include “and adult body mass” as parameters that were measured.

I did not read the Abstract, Simple Summary or Discussion, as these will be modified substantially in accordance with the revised statistical approaches.

Comments on the Quality of English Language

Overall, readability is good. However, the Introduction has some passive writing, e.g. the middle of the 3^rd sentence in the Introduction, which begins as “It may seem that….” A few subject-verb pairs are incorrect.

No need for writing "4th larval instar"--"4th instar" is sufficient. The larval stage is assumed.

Author Response

Language used by the authors throughout the Methods and Results sections indicates that they have misconceptions about their own experimental set up and thus about how proceed in analyzing and reporting the results.

The authors repeatedly misspeak about treatments and treatment levels, and reference them in erroneous phrases such as “among the six treatments” or “between the six treatments” and mixing these with statements like the “treatment (C or P).” There were only two treatment factors (pheromone and larval density)—one had two levels (Clean and Pheromone) and the other had three (1, 4 and 8 larvae) –for a total of five treatment levels. However, as this was a factorial design, there were six (2 x 3) treatment combinations.

= We apologize. In literature, „treatment“ is either a factor with several levels or a group defined by a combination of factor levels. But it should be kept consistent, which we did not keep during alternation of writing between the two authors. To avoid this confusing writing, we decided to avoid the word “treatment” altogether, and we use the words “factor”, “level”, and “combination” throughout.

The interpretation of the significance or lack of significance of each treatment and their interaction was not consistently handled appropriately by the authors. For the effect on 4^th instar development duration, pheromone, larval density, and their interaction were all significant. Thus, it is appropriate to compare means among the individual treatment combinations using Least Squares Means or something equivalent. However, this is tricky, given that the random factor of the dish was highly significant. Regardless, it appears that the authors erroneously performed the Tukey HSD for each main factor but also used these within-factor Tukey results to compare means across both treatment factors, although it was not clear from the description of stat procedures in the Methods section. Thus, Fig. 1 appears incorrect.

= We are not sure if we understand the reviewer. We compared all six combinations of two factors by the Tukey test, not the factors separately. We added a short text about this to the methods and figure legends.

For the effect on adult body mass, pheromone and larval density treatments were each significant; Their interaction was non-significant. Thus, differences revealed by a Tukey HSD test should have been confined to P vs. C and then separately among the 3 larval densities—all six treatment combinations should NOT have been compared against one another (e.g., “Body mass in treatment C8 was lower than in C1” is inappropriate). Thus, Fig. 2 is invalid.

= We disagree. It is allowed to compare separately the levels of each factor (which is not in the case of duration above), but we can still look at each of the six combinations separately and compare them. This is frequently done in multi-level ecological studies.

The highly significant random effect of dish needs some kind of explanation, although the authors do wisely note their caution in interpreting their results in light of the random effect. Could the random effect have arisen by the way the 4^th instars were obtained for placement into the Petri dishes? That is, were faster developing larvae the first to go into the dishes?

= High significance of the dishes ID represents a puzzle for us. The above proposed methodical arrangement was not responsible – we always distributed new larvae to both low- and high-density dishes. ?!

Based on the the attached Supplementary data file, could the dish effect have arisen because some of the dishes were maintained at room temperature whereas other were kept at 25 C?

= Neither did this make a difference among dishes. We propose at the end of the discussion to keep the time of the experiment short and the genetic source (parents) homogeneous.

The authors were incomplete in reporting their ANOVA results, supplying only the numerator degree of freedom and omitting the denominator d.f. What software was (and perhaps programs were) used to conduct the statistical analyses?

= We know we should report both, but our software Statistica 13 provided only one of the DF values after introducing the random effect model.

The title is inexact, mentioning only “larval development,” whereas measurements were also made on pupae and adults in the study.

= That is right. We changed the title.

The Introduction needs improvement. Studies cited in the Introduction are referred to inconsistently in either the past or the present tense.

= We are aware of this, and we want to keep the two tenses. In the beginning, we write some general and everlasting facts about the life of ladybirds, such as “pupae are immobile, and they are thus more prone to be cannibalized, “ which deserves the present tense. The citation in brackets is in this case simply one illustrating a generally known fact. But later, we give some concrete examples, such as „Adults and larvae of Coleomegilla maculata and Propylea quatuordecimpunctata attacked eggs, “ which was found in that one single study we cite, and thus we use the past tense in this situation. But there were two cases that did not keep this rule, and we corrected them.

In the first sentence of the last paragraph of the Introduction, “investigate” needs to be corrected. The last sentence of this paragraph needs to include “and adult body mass” as parameters that were measured.

= Thanks. Corrected.

I did not read the Abstract, Simple Summary or Discussion, as these will be modified substantially in accordance with the revised statistical approaches.

= They did not change substantially.

Comments on the Quality of English Language

Overall, readability is good. However, the Introduction has some passive writing, e.g. the middle of the 3^rd sentence in the Introduction, which begins as “It may seem that….” A few subject-verb pairs are incorrect.

= Some sentences were corrected.

No need for writing "4th larval instar"--"4th instar" is sufficient. The larval stage is assumed.

= Of course, thanks.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have responded constructively to the reviewer comments, and the manuscript has improved compared to the original version, particularly with respect to the statistical treatment of experimental non-independence. The use of mixed-effects models with dish identity as a random factor is appropriate and strengthens the robustness of the results. However, important issues remain regarding conceptual consistency, mechanistic interpretation, and data availability, which should be addressed before the manuscript can be accepted.

1. Although the authors have adopted the term “buffering” in parts of the Discussion and Conclusions, the Abstract and some sections still retain wording such as “development-enhancing pheromone” or similar implications. This inconsistency is potentially misleading, particularly in the absence of direct physiological or behavioral evidence. Please ensure consistent terminology throughout the manuscript (including the Abstract and Keywords), and avoid phrasing that implies direct developmental acceleration. A formulation emphasizing a potential buffering effect is more appropriate.
2. The authors note that GC-MS analyses of larval tracks are ongoing and will be published separately. While this is valuable future work, ongoing or unpublished studies cannot substitute for a clear statement of limitations in the current manuscript. Please include a dedicated Limitations section explicitly stating that no chemical identification or quantification was performed in this study, and that alternative sources of chemical cues cannot be excluded.
3. The introduction of mixed-effects models accounting for dish-level random effects is a substantial and commendable improvement. The revised analyses suggest that the main trends remain broadly consistent. Nevertheless, given the non-negligible contribution of random effects, interpretations should remain cautious, and strong generalizations should be avoided.
4. Adult body-mass data are currently provided via an external ResearchGate link. While data sharing is appreciated, external repositories do not fully replace journal-hosted supplementary materials. Please include the body-mass data as part of the official Supplementary Materials submitted with the manuscript to ensure long-term accessibility and reproducibility.
5. Although a 12-hour observation interval is common in related studies, a brief discussion of potential interval-censoring effects would strengthen the methodological transparency.

Overall, the manuscript has improved substantially compared to the original submission, particularly in terms of statistical rigor. However, several conceptual and data-related issues remain unresolved. I therefore recommend a minor (second-round) revision before acceptance.

Author Response

1. Although the authors have adopted the term “buffering” in parts of the Discussion and Conclusions, the Abstract and some sections still retain wording such as “development-enhancing pheromone” or similar implications. This inconsistency is potentially misleading, particularly in the absence of direct physiological or behavioral evidence. Please ensure consistent terminology throughout the manuscript (including the Abstract and Keywords), and avoid phrasing that implies direct developmental acceleration. A formulation emphasizing a potential buffering effect is more appropriate.

= You are right, we changed the expressions using “buffering” as a nice word to express the findings.

1. The authors note that GC-MS analyses of larval tracks are ongoing and will be published separately. While this is valuable future work, ongoing or unpublished studies cannot substitute for a clear statement of limitations in the current manuscript. Please include a dedicated Limitations section explicitly stating that no chemical identification or quantification was performed in this study, and that alternative sources of chemical cues cannot be excluded.

= Exactly, we have not yet published the other study with Nonacosanol as the major compound found in the track. Thus, we added limitations near the end of the discussion.

1. The introduction of mixed-effects models accounting for dish-level random effects is a substantial and commendable improvement. The revised analyses suggest that the main trends remain broadly consistent. Nevertheless, given the non-negligible contribution of random effects, interpretations should remain cautious, and strong generalizations should be avoided.

= We added a recommendation to make the results more homogeneous, decreasing the random effects.

1. Adult body-mass data are currently provided via an external ResearchGate link. While data sharing is appreciated, external repositories do not fully replace journal-hosted supplementary materials. Please include the body-mass data as part of the official Supplementary Materials submitted with the manuscript to ensure long-term accessibility and reproducibility.

= We are submitting Excel tables S1 and S2 with development time and body mass in cleaned, well-arranged form.

1. Although a 12-hour observation interval is common in related studies, a brief discussion of potential interval-censoring effects would strengthen the methodological transparency.

= Also added to the Limitations paragraph. In fact, a 12-hour interval is not common; many others still observe their insects only once daily!