How Much Does Fire Behavior of Leaf Litter Beds Change within Two Months?
Round 1
Reviewer 1 Report
This is an interesting manuscript that describes aspects of fire behaviour in three species in relation to attributes of their leaf litter. A manipulative experiment was performed to explicitly examine the effects of increasing time on fire behaviour. The manuscript presents new data and interpretations that will be of broad interest to flammability ecologists. I provide the following comments to assist the authors in revising their manuscript.
(1) Throughout the manuscript, the grammar needs attention in regard to several issues. The word ‘the’ is used where it shouldn’t be and it makes the writing appear clunky (e.g., in the first sentence of the Abstract, ‘the’ is not needed before ‘vegetation fuels’). There are instances throughout where plural/singular are mixed up (e.g., second sentence of the Abstract should be ‘are’ not ‘is’). I recommend that the manuscript is carefully edited to correct these linguistic errors in the manuscript.
(2) In the Abstract, the word ‘behaviour’ is used for all the fire attributes measured in the experiment. I know that one is word limited in the Abstract, but there needs to be a clearer description of what is meant by behaviour in the context of this study, so that the reader knows exactly what was measured about fire right from the start.
(3) There appears to be missing text in the fourth paragraph of the Introduction (L 68). This needs to be corrected.
(4) Add Krix et al. (2019) as a more recent and highly relevant reference in the area of relationships between functional traits and flammability in L 37 of the Introduction: Krix DW, Phillips ML, Murray BR (2019) Relationships among leaf flammability attributes and identifying low-leaf-flammability species at the wildland-urban interface. International Journal of Wildland Fire 28: 295–307.
(5) I suspect that the figure legends are too long, and some detail should be moved into text of the Materials and Methods section.
(6) The Results section needs the addition of F, d.f. and P values in addition to effect sizes for the ANOVA results.
(7) Some of the tables and figures in the supplementary material could be usefully moved into the manuscript text. There is important information presented in S1, S5, S6, S7 and S8, that really help to show the excellent work in done in this experiment.
Author Response
(1) Throughout the manuscript, the grammar needs attention in regard to several issues. The word ‘the’ is used where it shouldn’t be and it makes the writing appear clunky (e.g., in the first sentence of the Abstract, ‘the’ is not needed before ‘vegetation fuels’). There are instances throughout where plural/singular are mixed up (e.g., second sentence of the Abstract should be ‘are’ not ‘is’). I recommend that the manuscript is carefully edited to correct these linguistic errors in the manuscript.
(1) In order to improve our manuscript extensive language editing was performed.
(2) In the Abstract, the word ‘behaviour’ is used for all the fire attributes measured in the experiment. I know that one is word limited in the Abstract, but there needs to be a clearer description of what is meant by behaviour in the context of this study, so that the reader knows exactly what was measured about fire right from the start.
(2) We made several attempts to mention individual, measured variables in the Abstract.
Nevertheless, we found it impossible to do so, without exceeding the limit of 200 words setby the Journal.
(3) There appears to be missing text in the fourth paragraph of the Introduction (L 68). This needs to be corrected.
(3) This paragraph starts with a reference. We added the authors name to make it clearer.
(4) Add Krix et al. (2019) as a more recent and highly relevant reference in the area of relationships between functional traits and flammability in L 37 of the Introduction: Krix DW, Phillips ML, Murray BR (2019) Relationships among leaf flammability attributes and identifying low-leaf-flammability species at the wildland-urban interface. International Journal of Wildland Fire 28: 295–307.
(4) There are numerous, recently published works dealing with relationships between functional traits and flammability. We agree that the list of references provided here could be much longer.
Opposite to the suggestion given by the reviewer, we chose not to include the work of Krix et al. (2019) in the manuscript. The reason for this is that our Institute does not have direct access to the International Journal of Wildland Fire. We are able to access those articles through the inter-library exchange. We started the process of getting access to the suggested article as soon as we got review report, but as yet we did not receive the article in question. Due to the very short resubmission deadline, we are not able to wait for it. As a basic rule, we do not refer to any work we did not read in detail, thus we are currently unable to comply with the reviewer’s suggestion.
(5) I suspect that the figure legends are too long, and some detail should be moved into text of the Materials and Methods section.
(5) The statistical analyses performed on our data have been explained in detail in the “Material and Methods” section of the original manuscript. When originally writing the figure legends, we followed the suggestion that all figures should stand alone and provide all the information on performed statistical analyses. Since both reviewers found our legends to detailed, we tried to shorten them.
(6) The Results section needs the addition of F, d.f. and P values in addition to effect sizes for the ANOVA results.
(6) We added d.f. and F-values for all ANOVAs. P-values are added if P > 0.001. Lower P-values were far below the 0.001 threshold. Giving their exact value would have resulted in very long numbers (10-x notation) needing a lot of space in the graphs.
(7) Some of the tables and figures in the supplementary material could be usefully moved into the manuscript text. There is important information presented in S1, S5, S6, S7 and S8, that really help to show the excellent work in done in this experiment.
(7) We thank the reviewer for complimenting our work. Nevertheless, we decided not to move supporting material into the manuscript text. During the course of the experiment a large amount of data was generated. Data included in the manuscript text was carefully chosen in order to address the main issues we were interested in (i.e. effects of the exposure duration on structure and fire behaviour of leaf litter beds) without losing readers attention. We do consider the information provided in the supplementary material important in proving the quality of our work, corroborating the claims made in the main text and provide the measured values for the interested reader, but we believe this information is not mandatory for following the argumentation in the manuscript text.
The reason for not including S1 supporting information into the main article is purely practical. Photos were taken for documentation purposes only and their quality does not comply with the publishing requirements (too low resolution of the original photos).
The reason for not including S5 in the main article is already given in the text (L214- 219 of the original text). The information provided in the S6 is almost identical to the information given in Figure 3. Thus, to avoid redundancy, we chose to use one pair of graphs in the main article and another one in the supporting information (as explained in the L242- 244).
One of the early versions of the manuscript did include the information which is now in S7 and S8. This version of the manuscript was quite long and very repetitive. Our impression was that such a text is tiresome to read and includes too much information which is not essential for presenting the most important findings of our work – two species which exhibited very different behaviour when tested as freshly prepared samples grouped together after two months exposure to outside weather conditions. Thus we chose to use PCA scores in the main article and provide the raw data (S9) as well as the results of the analysis performed on the individual measured parameter in the supporting material (S7, S8).
Two alterations were made, which were not related to the reviewers’ comments.
We realised, we made an error in the Figure 2. Instead of bulk density and packing ration, originally submitted Figure 2c and 2d presented sample height and bulk density, respectively. We sincerely apologise for this error.
Due to administrative reasons, in Acknowledgments, “National Park Mljet” was replaced with “Mljet National Park Public Institution”. The first term designates protected area, whereas the second designates the institution governing it.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This was an interesting manuscript- thanks for doing this important research!
My comments below are intended to help make this good ms even better- I hope they help.
Title is a little cheeky, but the Keywords capture the terms I would have used in the title.
Abstract
L10: "Fire behavior is well-recognized as a function of fuel characteristics, but in practice the dynamics of fuels are often overlooked" would be a better way to begin the ms.
L17: Delete "one"
L19: Reword to "be attributed to pore space within the fuel bed, with more pore space relating to more dramatic treatment effects."
L23: Reword to "an issue that should be addressed in future research."
Keywords are very good (might make them alphabetical)
Introduction- this section is unnecessarily long, but well done and documented.
L27: Delete first sentence and get to the point more quickly (and renumber citations).
L33: Also add intensity or flame height or another metric of energy flux.
L38: Reword to "Even though substantial work has been devoted to leaf litter.."
L53: Delete "Furthermore,"
L55: Delete unnecessary "the" before "BD"
L57-67 could have a more active voice (We did X or we tested) rather than passive voice.
L64: Insert "the" before "first"
L68: Insert "Ormeno and others [21]" so the paragraph doesn't begin with a numeral.
L68-76 could be condensed and combined.
L77-91 is appreciated, but could use condensing and abbreviation.
L92-96: I would delete this text and replace with "The implications of this research are X and Y.. "
L97-100 could be moved below into the Methods.
Materials and Methods- this section is well done, if long. Try to abbreviate where you can without losing the necessary detail.
L112: Should be "weighed" and "weighing"
L113: Reorganize sentence that begins "Furthermore," with "To characterize fuel particles, we removed 50..."
L118-119: I wish you had combined this paper with the other one you may have.
L131: remove superscript "2"
L137: About how long and what rate of watering took place?
L144 area: The climate of Stuttgart comment and its relevance to Croatia is offputting and will invite criticism. I am fine with just saying the fuelbeds were weathered at the Lab site and not bringing it back to the Mediterranean relevance. Consider revising.
L156: date formatting should be 15 March 2017 for all of these (incl. L168).
L161: Add or replace mass change with %change +-std error.
L180: Remove superscript
L181: Ditto.
L182: "mash" should be "mesh"
L196: Remove ":"
L211: Make "-1" superscripted.
L223: "by the average particle density" is repeated.
L236: "factor" should be "factors"
L239: "extend" should be "extent"
L246: Combine text with previous paragraph.
Results- a general comment is to spell out terms when possible and favor the actual measured variables (ie flame height) rather than the PC scores. This syntax helps the reader understand what happened without flipping back or trying to recall what PC1 was actually related to.
L262: What was PC3 related to and how much additional % variance was explained by PC3?
L278: See general comment on PC vs. measured variables.
L284: make "point" singular
L286-314: See general comment.
Figure 2: You have space, so I would spell out as much within the figures as you can afford- spec should be "species" and "treatment on x-axis could be elaborated as "litter treatment" or something.
Fig 2: It would be helpful to label the axes for PC1 and 2 with the terms they are related to so the reader can infer how treatment and BD were/weren't related to the flammability terms.
Fig 2: Spell out y-axis labels on c and d.
L318-33 is long- see if you can add to the graphs to remove wordy text here or remove redundant text sufficiently to keep the story intact.
L353-354: Superscript the "2" in R2.
Fig 3: spell out bulk density on x-axes. Could also add the flammability terms on the y-axis to inform what PC1 and PC2 are.
Discussion- section could be refined and abbreviated, but is well done.
L376: Spell out FH, RoS, and FD etc in this paragraph.
L377: You didn't measure terpene content, so I would avoid talking about it unless presented as a potential additional explanatory variable (beyond BD as presented in Fig 3a).
L391-420 could be abbreviated considerably- try two paragraphs with distinct ideas.
L428: "insure" should be "ensure"
L431-432 and later: L, F, and H "layers" are better interchanged as Oi, Oe, and Oa horizons.
L438: italicize P. ponderosa
L455: "premises" should be "premise"
L456: Reword to "treatment effects on fire behavior"
L458: Reword to "supports this argument"
L474-486: section is long and wandering- tighten into a shorter single paragraph.
L504: This is a major point (well done!!!), but could be improved- suggest deleting "We imply that,"
L509: Insert "."
Conclusions- your last paragraph of the Discussion section (L500-509) is the best section of the ms and would be best as a concluding statement. I would either delete the "Conclusions" section or cut some of 500-509 and paste judiciously here.
Author Response
L10: "Fire behavior is well-recognized as a function of fuel characteristics, but in practice the dynamics of fuels are often overlooked" would be a better way to begin the ms.
L10: Altered accordingly.
L17: Delete "one"
L17: Altered accordingly.
L19: Reword to "be attributed to pore space within the fuel bed, with more pore space relating to more dramatic treatment effects."
L19: Altered accordingly.
L23: Reword to "an issue that should be addressed in future research."
L23: Altered accordingly.
Keywords are very good (might make them alphabetical)
Keywords are in the alphabetical order now.
Introduction- this section is unnecessarily long, but well done and documented.
L27: Delete first sentence and get to the point more quickly (and renumber citations).
L27: We decided not to delete the first sentence of the introduction. It might not be absolutely necessary for this manuscript, but, in our opinion, it does not deteriorate the quality of the revised manuscript.
L33: Also add intensity or flame height or another metric of energy flux.
L33: As many different variables can be used to describe fire, we deliberately chose not to include individual parameters but only mention some broad aspects of fire behaviour. Flame height and energy flux are broadly related to the ability of vegetation to support fire and facilitate fire spread. Thus, we choose not to add these or any other individual metrics. In our opinion including all, potentially relevant metrics could lead to an unnecessary long list of variables which were not measured in our work.
L38: Reword to "Even though substantial work has been devoted to leaf litter.."
L38: Altered accordingly.
L53: Delete "Furthermore,"
L53: Altered accordingly.
L55: Delete unnecessary "the" before "BD"
L55: Altered accordingly.
L57-67 could have a more active voice (We did X or we tested) rather than passive voice.
L55: Altered accordingly.
L64: Insert "the" before "first"
L55: Altered accordingly.
L68: Insert "Ormeno and others [21]" so the paragraph doesn't begin with a numeral.
L55: Altered accordingly.
L68-76 could be condensed and combined.
L68-76: We do realise that some sections could use shortening/condensing. Nevertheless, at these points we considered it more important do discuss the given issues in detail, rather than to reduce the text volume.
L77-91 is appreciated, but could use condensing and abbreviation.
L77-91: same as L68-76. Thus, we chose to keep original, detailed hypotheses.
L92-96: I would delete this text and replace with "The implications of this research are X and Y.. "
L92-96: We chose to leave this section intact.
L97-100: could be moved below into the Methods.
L97-100: In our opinion, this statement is important enough to be retained in the Introduction as the choice to focus on the individual parameters and not on the flammability note shapes the whole manuscript.
Materials and Methods- this section is well done, if long. Try to abbreviate where you can without losing the necessary detail.
L112: Should be "weighed" and "weighing"
L112: Altered accordingly.
L113: Reorganize sentence that begins "Furthermore," with "To characterize fuel particles, we removed 50..."
L113: Altered accordingly.
L118-119: I wish you had combined this paper with the other one you may have.
L118-119: We wished it too. Our initial intention was to combine all 115 tests into one single paper. However, we found it difficult to introduce the effects of aging on the leaf litter material and then transfer these findings onto mixtures in one single paper. Thus, we decided to focus on the effects of aging on monospecific leaf litters in the first paper, and on the mixtures in the second paper. As the second paper is currently still being prepared, we were not able to submit the first and the second paper simultaneously.
L131: remove superscript "2"
L131: Altered accordingly.
L137: About how long and what rate of watering took place?
L137: We did not track the total time or watering rate. We made sure that the whole area was evenly watered and that watering rates were not exaggerated in comparison to natural precipitation. Watering was performed in such a way that the person performing the watering steadily moved along the length of the experimental area, first watering the area close to one edge, then the middle part, then the area next to the second part. When the watering in the longer direction was finished, no dripping was observed at the lower side of any of the samples. Thus it was repeated by watering “in stripes” perpendicular to the longer side of the experimental area. This time, the person performing the watering counted the seconds until dripping was observed on the lower side of all samples in the first “stripe” and repeated it for each following “stripe”. As we don’t have the information on the duration of the “longitudinal” watering, this provides only a partial information. During both runs, we made sure that the edges of “stripes” overlapped in order to obtain homogenous watering of the whole area. Possible inhomogeneities in watering were addressed through the random block design of the experiment. We do realise that providing exact information on the water amounts and rates would improve repeatability of our experiment. Nevertheless, as samples were well drained (positioned on the mesh surface) and exposed in the greenhouse for a longer period of time (fluctuations in the air humidity and temperature) we do not consider water amounts and rates crucial for the observed effects of the greenhouse treatment.
L144 area: The climate of Stuttgart comment and its relevance to Croatia is offputting and will invite criticism. I am fine with just saying the fuelbeds were weathered at the Lab site and not bringing it back to the Mediterranean relevance. Consider revising.
L144: We do agree that the climate of Stuttgart and the island of Mljet differ greatly. If a longer exposure period had been used, we would also agree that commenting relevance of Stuttgart climate for Croatian island is offputting. Nevertheless, duration and time frame of the exposure period were carefully chosen. Our initial intention was simply to prevent exposing litter under conditions which would rarely occur in the place of origin (e.g. temperatures below 0°C, snowfall). After carefully examining and comparing climatic data between the two locations (Stuttgart, Mljet) we realised that the average temperature during the period mid-March to mid-May in Stuttgart corresponds well to late autumn/early winter temperatures on the island of Mljet. Furthermore, very often a decrease in the average daily temperature occurs in late April/early May in Stuttgart. Thus, by choosing mid-March for the beginning of the exposure period and mid-May for its end, we were hoping that the exposed material would experience a decrease in the average daily temperature towards the end of the exposure. This was the case in the year of the study, as shown in S2. Thus, we consider our claim that “spring weather conditions in Stuttgart roughly correspond to autumn/winter weather condition on the island of Mljet” to be justified.
L156: date formatting should be 15 March 2017 for all of these (incl. L168).
L156: Altered accordingly.
L161: Add or replace mass change with %change +-std error.
L161: changed with %change +-std error
L180: Remove superscript
L180: Altered accordingly.
L181: Ditto.
L 181: Altered accordingly.
L182: "mash" should be "mesh"
L182: Altered accordingly.
L196: Remove ":"
L196: Altered accordingly.
Accepted and altered accordingly.
L211: Make "-1" superscripted.
L211: Altered accordingly.
L223: "by the average particle density" is repeated.
L223: Repeated section was removed.
L236: "factor" should be "factors"
L236: Altered accordingly.
L239: "extend" should be "extent"
L239: Altered accordingly.
L246: Combine text with previous paragraph.
L246: Altered accordingly.
Results- a general comment is to spell out terms when possible and favor the actual measured variables (ie flame height) rather than the PC scores. This syntax helps the reader understand what happened without flipping back or trying to recall what PC1 was actually related to.
Results- a general comment
We chose not to replace PC-scores with the actual measured variables. Our interest lies in the overall change of fire behaviour. One of the early manuscript versions was based on the actual measured parameters and included PCA only in order to examine the fire behaviour space defined by the tested samples and to determine the position of individual samples within it. Nevertheless, this version of the manuscript was very long (since one should at least shortly describe all presented data) and repetitive (parameters loading highly on a single PC do exhibit similar behaviour). The decision to focus on PC-scores in the submitted manuscript is based on the fact that we focus on the change in fire behaviour over time.
Thus, we considered PC-scores sufficient to present the general pattern. Nevertheless, as we realise that the measured variables are important for assessing the quality of our work and comparing our work with other studies, we provide not only raw data (S9), but also ANOVA results (S7) and regression analyses (S8) for the actually measured variables in the supporting materials.
In the Figure 2, we chose not to label PC1 and 2 with the terms they are related to as this might create the false impression that individual, actually measured variables exhibit the same relationship with bulk density as PC scores. As we included a relatively low number of variables, this would not lead to large deception, but could still lead to false interpretation. For instance, the relationship between bulk density and the PC1 score is significant within Q. pubescens, but only one of the actual measured variables loading highly onto PC1 exhibits a significant relationship with BD within the given species (S8). If a reader is interested into relationship between the actually measured parameter and bulk density, the exact values are provided in supporting material S8.
For the same reason we did not include the actually measured variable on the y-axes of Fig. 3.
L262: What was PC3 related to and how much additional % variance was explained by PC3?
Before deciding to include only the first two PCs in the manuscript we did take into account all 5 PCs.
R-results for all PC-s are given:
Standard deviations (1, .., p=5):
[1] 1.6432083 1.0565872 0.8104246 0.5810727 0.4348063
Rotation (n x k) = (5 x 5):
PC1 PC2 PC3 PC4 PC5
ROS 0.5462250 -0.04504747 0.12282418 0.5960848 -0.57376480
FD -0.4479217 -0.14941833 0.74371859 0.4319418 0.19325942
SD 0.4007249 0.45024936 0.62651382 -0.4931021 -0.03202754
FH 0.5589037 -0.13837619 -0.03181882 0.2366314 0.78196665
UNCONS -0.1674830 0.86820068 -0.19561411 0.3987042 0.14473128
> summary(burn.PCA)
Importance of components:
PC1 PC2 PC3 PC4 PC5
Standard deviation 1.643 1.0566 0.8104 0.58107 0.43481
Proportion of Variance 0.540 0.2233 0.1314 0.06753 0.03781
Cumulative Proportion 0.540 0.7633 0.8947 0.96219 1.00000
The decision to include only the first two PCs into the manuscript was based on the Scree plot and the results of the “vvs”-R function. One could argue that we should have included PC3 as it reveals a positive relationship between FD and SD. Nevertheless, it explains only 13.14% of the total variance and was excluded by “vvs”. Furthermore, as we included only 5 variables in the PCA, we considered it exaggerated to extract 3 PCs. We considered it unnecessary to explain in details why we chose to present only the first two PCs, as this type of explanations is rarely found in the literature. Nevertheless, if considered necessary, we could include the argument given above into the manuscript.
L278: See general comment on PC vs. measured variables.
L278: See response to general comments.
L284: make "point" singular
L274: Altered accordingly.
L286-314: See general comment.
L286-314: See response to general comments.
Figure 2: You have space, so I would spell out as much within the figures as you can afford- spec should be "species" and "treatment on x-axis could be elaborated as "litter treatment" or something.
Figure 2: Altered accordingly.
Fig 2: It would be helpful to label the axes for PC1 and 2 with the terms they are related to so the reader can infer how treatment and BD were/weren't related to the flammability terms.
Fig 2. See response to general comments.
Furthermore, we realised, we made an error in the Figure 2. Instead of bulk density and packing ration, originally submitted Figure 2c and 2d presented sample height and bulk density, respectively. We sincerely apologise for this error and corrected it in resubmission. In text data was correct.
Fig 2: Spell out y-axis labels on c and d.
Fig 2: Altered accordingly.
L318-33 is long- see if you can add to the graphs to remove wordy text here or remove redundant text sufficiently to keep the story intact.
L318-333: We made an effort to make figure legends shorter.
L353-354: Superscript the "2" in R2.
L353-354: Altered accordingly.
Fig 3: spell out bulk density on x-axes. Could also add the flammability terms on the y-axis to inform what PC1 and PC2 are.
Fig 3: See response to general comments.
Discussion- section could be refined and abbreviated, but is well done.
L376: Spell out FH, RoS, and FD etc in this paragraph.
L376: We chose to leave the abbreviations, as this is in accordance with the Instructions for Authors of the Journal: “Abbreviations should be defined in parentheses the first time they appear in the abstract, main text, and in figure or table captions and used consistently thereafter.”. We could, if considered necessary, remove the abbreviations completely.
L377: You didn't measure terpene content, so I would avoid talking about it unless presented as a potential additional explanatory variable (beyond BD as presented in Fig 3a).
L377: Part of the sentence dealing with terpene content was deleted.
L391-420 could be abbreviated considerably- try two paragraphs with distinct ideas.
L391-420. See L68-76.
L428: "insure" should be "ensure"
L428: Altered accordingly.
L431-432 and later: L, F, and H "layers" are better interchanged as Oi, Oe, and Oa horizons.
L431-432: Altered accordingly.
L438: italicize P. ponderosa
L438: Altered accordingly.
L455: "premises" should be "premise"
L455: Altered accordingly.
L456: Reword to "treatment effects on fire behavior"
L456: Altered accordingly.
L458: Reword to "supports this argument"
L458: Altered accordingly.
L474-486: section is long and wandering- tighten into a shorter single paragraph.
L474-486: See L68-76.
L504: This is a major point (well done!!!), but could be improved- suggest deleting "We imply that,"
L504: Altered accordingly.
L509: Insert "."
L509: Altered accordingly.
Conclusions- your last paragraph of the Discussion section (L500-509) is the best section of the ms and would be best as a concluding statement. I would either delete the "Conclusions" section or cut some of 500-509 and paste judiciously here.
Conclusions: As Fire does not insist on Conclusions section and we agree with the reviewer that the major point of our work is stated in the L 500-509 of the original manuscript, we decided to delete the “Conclusions”.
