Characterization of Wildland Fuels Based on Topography and Forest Attributes in North-Central Appalachia

Round 1

Reviewer 1 Report

This study measures fall fuel and vegetation conditions of a 58ha watershed in southeastern Ohio using 94 sample plots. The paper reports useful insights into this region’s pattern of fuel variation by aspect and overstory composition with possible implications for understanding fire behavior.

 However, I was confused by what this paper was actually about from the title through the methods section as I was expecting more direct analysis of a fire.  It would have helped if in the title, “…fire environment…” would have said fuel / veg attributes, as this descriptive paper is largely about fuel characterization, not insights from an actual burn. Also, having fall fuel data is important for research, but some fall fuels change rapidly during October in eastern deciduous forests, so the value of the unspecified “October” timing of data collection could affect fuel compaction, fine fuel availability, fire duration, rate of spread, patchiness, and intensity.

ABSTRACT--The abstract needs to be substantially rewritten to better justify, frame and summarize this research project.

Line 9: the abstract’s topic sentence is awkward and true by definition. Rewrite it. The concept of a “fire environment” is open ended. It is unnecessary and confusing jargon.

Line 11, 13: one could argue that “a comprehensive description” is never achievable. Be more specific without overstatement.

Line 14: If you reference a “study site” in the Abstract, you should first introduce the system briefly. What continent? vegetation type? fire regime?

Line 16: “Ground layers” seems non-standard terminology. Do you mean the herbaceous veg or the surface fuels?  How do these components relate to fuels as dead herbaceous cover and shrubs are also fuels?

Line 18-19: I don’t think this statement is warranted based on the nature of this descriptive study. You didn’t observe this from fire effects. There are other explanations unconsidered such as spread direction and fuel moisture which can vary month to month and event to event.

INTRODUCTION

Line 30: Don’t you mean local and landscape scales, not regional?

Line 31: Change “structure and compositions” to “structure and composition” or “structures and compositions”.

Line 70: This is somewhat misleading. Most eastern oak forests have not experienced “centuries of fire exclusion”. A major additional relevant factor is widespread logging a century ago and tree species loss.

MATERIALS AND METHODS--Critical information is missing at the necessary time in this read. That is, until line 140, I was thinking this was an experimental burn study; not a simple fuel characterization study. That problem needs to be fixed in the title, abstract or introduction. Having 94 is impressive, but the number of them that fall in the various topographic groups is seemingly variable and not stated. It’s important to report those N’s by aspect and topo position somewhere in this section.

Line 93: How representative of this area’s landscape is this particular small watershed? To present results as meaningful at scale for managers, this point is important.

Figure 1: What explains the areas without sample points on the grid? Is this related to accessibility or forest edge constraints as mentioned on line 100?  From recent aerial photos of this watershed, I see that “forest edge” apparently means harvesting blocks. Could those recently formed edges affect woody debris or fuels? They’d be easy to include in Fig 1.

Line 99: This grid design gives great spatial distribution, but does it capture less common vegetation types as well as a stratified approach? That is, there are few lower slope plots and possible redundancy in the upper slopes.

Line 102, Figure 2: The reference is to just aspect in the text, but the Figure 2 reference is to sub-maps A, B and C, with B (slope) described later on and C (elevation) never mentioned in the text but is apparently a place holder for slope position.  This could be tightened up.

Line 104-6, Figure 2: Comment: There are simple and more robust ways to calculate slope position in GIS than through visual estimation. That would have given you a map similar to Fig 2 A and B instead of C (elevation).

Line 106: You state how many total plots, but it is not clear how many plots fall into the various divisions—how many S, W, N, E plots do you have? How many by slope position? What are your N’s? There seems to be a large representation of west and south aspects and relatively few N and E aspects from Figure 2.

Line 114: Change “Forest attributes data” to “Forest attribute data”.

Line 124: Do you mean “recorded trees”, not “recording trees” here?

Line 127: add the missing word “were” between “as” and “ground”.

Line 139: Was leaf fall actually completed here by October of 2022? Oak leaf retention typically goes well beyond then in nearby landscapes, while mesophytic species tend to fall quicker. This statement implies that there were no leaves left on the trees when sampling occurred. Was that true?

Line 160: Multiple tests require a standard Bonferroni correction, and when used should be stated in the Methods.  This could impact your characterization of differences found for most if not all analyses.

RESULTS

There is no reason to say “significant difference” in this section or in the figure captions (anywhere) when defining difference with a test of significance. Just say “difference”. 

Line 179: You may need to change “by aspects” to “by aspect”.

Captions for Figure 2, Figure 3, Figure 4, Figure 5, Figure 6: Restating the degree divisions of aspects here in the figures is excessive. Such detail belongs in the methods, and only need to be provided once.

Line 296: “log-transformed” belongs in methods, not results.

Figure 10: This is a nice figure but see comment in Line 160 above for the reported value’s relevance.

DISCUSSION

Line 352: To be fair, your results do not show that fuel composition influences fire temperature because you conducted no burn; you’re saying there is implications for fire temperature based on understanding from other research. Don’t overstate your actual results; distinguish those better from suggested implications.

Line 360: Change “compositions” to “composition”.

Line 351-361, 365: There seems to be a lot of emphasis on spread /intensity implications, but not as much about duration. There could be more discussion of that here as you measured fuel bulk density and depth (lines 141-142) as well as associated tree sources.  The duff qualities could be different and that could affect fire outcomes in a number of ways beyond what is discussed. See, for example, Carpenter et al 2021 Ecosystems 24:1059-1074. 10.1007/s10021-020-00568-7

Line 376, 381: How much of this expected difference in severity would results from simple sun exposure affecting fuel moisture and not differences in fuels. Aspect differences in fuel moisture are enhanced in the leaf-off season when sun angles are low.

Line 391: It seems to be a bit of an overstatement to declare that the simple presence of more oak spp. litter affects fire spread. In drought, when these landscapes burn, litter is normally not particularly limiting—that is, you’ve not shown that fuel is limited, and you haven’t made any direct burn observations. In this entire paragraph, there’s lots of inference. Instead, focus on what this paper actually contributes. You’re not demonstrating fire spread potential; you’re just inferring it from others’ work and your plot data that shows there are some differences in fuel bed properties with certain implications. Avoid overstatement with more careful wording.

Line 394-397: If litter fuel drives fire spread, and it isn’t limiting at this time of year, there is arguably not a “higher level of potential fire”. You might move past this by focusing on potential fire intensity rather than spread potential. That is, what do mean by “potential fire”? The implications of fire on regen or mortality aren’t from just any fire, but only certain types of fire when fuel or fuel moisture may not be limiting.

CONCLUSIONS

Line 423: Rewrite: This topic sentence’s emphasis on species distributions is not the focus of this paper and it has been understood for decades. Instead, focus on this paper’s better contribution to fuel patterns and its implications for fire. Fuel is lost even more in sentence two that suggests that the documented species patterns can have consequences for fuel and fire, but fuel has been directly described, not just inferred.

Line 428: As mentioned above, this expectation for more fire on south aspects also follows from our understanding of fuel moisture, not just fuel type.

Line 429a: What does “lowest fire” mean? Intensity? Frequency?

Line 429b, 430: Wording. Composition “is”, not “are”. Also “data show” not “data shows”.

Line 429, 430: Nuance the phasing better. Your patterns have implications for subsequent fire intensity. You haven’t demonstrated that because you did not conduct burn observations. This is inferred.

Line 431: Confusing statement. Isn’t overstory tree species composition a forest attribute difference? The overstory attributes cause the differences in understory fuel examined.

REFERENCES

Line 481, Line 524: Fix cap letters to standardize formatting!

Line 469, 486, 508, 532, etc.: Fix Use Of Title Caps When elsewhere they are not caps. Standardize formatting!

Occasional grammar issues. Nothing major.

Author Response

Reviewer 1  

ABSTRACT--The abstract needs to be substantially rewritten to better justify, frame and summarize this research project.

1. Line 9:the abstract’s topic sentence is awkward and true by definition. Rewrite it. The concept of a “fire environment” is open-ended. It is unnecessary and confusing jargon.

Changed to “has the potential to subsequently influence variations in fire behavior”

2. Line 11, 13:one could argue that “a comprehensive description” is never achievable. Be more specific without overstatement.

Changed to “a fine-scale description of fuel patterns and their relationship with overstory and understory attributes”.

3. Line 14: If you reference a “study site” in the Abstract, you should first introduce the system briefly. What continent? vegetation type? fire regime?

Changed to “within an oak-hickory forest located in southeast Ohio, USA, which historically fell within fire regime group I with a fire return interval ranging from 7 – 26 years”.

4. Line 16:“Ground layers” seems non-standard terminology. Do you mean the herbaceous veg or the surface fuels?  How do these components relate to fuels as dead herbaceous cover and shrubs are also fuels?

Changed to “shrub and herbaceous layers, surface fuels”.

5. Line 18-19: I don’t think this statement is warranted based on the nature of this descriptive study. You didn’t observe this from fire effects. There are other explanations unconsidered such as spread direction and fuel moisture which can vary month to month and event to event.

I am not sure which sentence it is talking about…in Lines 18-19, “Some of the results show that the oak fuel load was highest on south-facing slopes and on upper slope positions while maple fuel loads were similar across all aspects and slope positions.” The sentence did not talk about fire.

INTRODUCTION

6. Line 30: Don’t you mean local and landscape scales, not regional?

Changed.

7. Line 31: Change “structure and compositions” to “structure and composition” or “structures and compositions”.

Changed.

8. Line 70:This is somewhat misleading. Most eastern oak forests have not experienced “centuries of fire exclusion”. A major additional relevant factor is widespread logging a century ago and tree species loss.

If you mean line 80, “centuries” have been changed to “decades” for more clarity.

MATERIALS AND METHODS--Critical information is missing at the necessary time in this read. That is, until line 140, I was thinking this was an experimental burn study; not a simple fuel characterization study. That problem needs to be fixed in the title, abstract or introduction. Having 94 is impressive, but the number of them that fall in the various topographic groups is seemingly variable and not stated. It’s important to report those N’s by aspect and topo position somewhere in this section.

The title was changed to better reflect the nature of the study. It was changed to “Characterization of wildland fuels based on topography and forest attributes in North-Central Appalachia”

9. Line 93: How representative of this area’s landscape is this particular small watershed? To present results as meaningful at scale for managers, this point is important.

A statement was added to emphasize the importance: “An area of 58 hectares located within an oak-hickory forest, referred to as Morgan Hollow, was selected for this study as this forest type dominates this region and accounts for nearly half of the forest cover in the eastern United States.”

10. Figure 1:What explains the areas without sample points on the grid? Is this related to accessibility or forest edge constraints as mentioned on line 100?  From recent aerial photos of this watershed, I see that “forest edge” apparently means harvesting blocks. Could those recently formed edges affect woody debris or fuels? They’d be easy to include in Fig 1.

An explanation regarding missing points on the grid was added in the caption of Fig 1.

11. Line 99: This grid design gives great spatial distribution, but does it capture less common vegetation types as well as a stratified approach? That is, there are few lower slope plots and possible redundancy in the upper slopes.

Added in the caption of Fig 2. There are indeed few lower slope plots compared to upper and middle slope plots, but a total of 21 lower slope plots still provide sufficient data, especially when the plots are established by a grid work, and 60m is a good plot interval to capture less common vegetation types. 

12. Line 102, Figure 2: The reference is to just aspect in the text, but the Figure 2 reference is to sub-maps A, B and C, with B (slope) described later on and C (elevation) never mentioned in the text but is apparently a place holder for slope position.  This could be tightened up.

Changed the elevation map to the slope position map.

13. Line 104-6, Figure 2:Comment: There are simple and more robust ways to calculate slope position in GIS than through visual estimation. That would have given you a map similar to Fig 2 A and B instead of C (elevation).

Changed the elevation map to the slope position map.

14. Line 106: You state how many total plots, but it is not clear how many plots fall into the various divisions—how many S, W, N, E plots do you have? How many by slope position? What are your N’s? There seems to be a large representation of west and south aspects and relatively few N and E aspects from Figure 2.

Added in the caption of Fig 2.

15. Line 114:Change “Forest attributes data” to “Forest attribute data”.

Changed.

16. Line 124: Do you mean “recorded trees”, not “recording trees” here?

Changed to “recorded trees”.

17. Line 127:add the missing word “were” between “as” and “ground”.

Changed to “and the height of the tallest shrub was measured. ”

18. Line 139: Was leaf fall actually completed here by October of 2022? Oak leaf retention typically goes well beyond then in nearby landscapes, while mesophytic species tend to fall quicker. This statement implies that there were no leaves left on the trees when sampling occurred. Was that true?

Added the specific fuel collection period, “Litter samples, which include 1-hour (0 – 0.6cm diameter) and 10-hour (0.6 – 2.5cm diameter) fuels [30] were collected after leaf fall from October 30th to November 5th, 2022, and during the period of the opening of the prescribed burn window.”

19. Line 160:Multiple tests require a standard Bonferroni correction, and when used should be stated in the Methods.  This could impact your characterization of differences found for most if not all analyses.

Added the standard Bonferroni correction in the method “using two-way ANOVA followed by a standard Bonferroni correction and a post-hoc Tukey-Kramer test (significance level ∝ = 0.05).”, the corresponding p-value and differences letters in the results and figures were changed.

RESULTS

20. There is no reason to say “significant difference” in this section or in the figure captions (anywhere) when defining difference with a test of significance. Just say “difference”. 

Removed “significant”

21. Line 179:You may need to change “by aspects” to “by aspect”.

Changed.

22. Captions for Figure 2, Figure 3, Figure 4, Figure 5, Figure 6:Restating the degree divisions of aspects here in the figures is excessive. Such detail belongs in the methods, and only need to be provided once.

Changed.

23. Line 296: “log-transformed” belongs in methods, not results.

Changed to “The data set was log-transformed to reduce the effect of outliers in the data when doing correction analyses.” In the method part.

24. Figure 10:This is a nice figure but see comment in Line 160 above for the reported value’s relevance.

A standard Bonferroni correction was added when conducting correlation analysis in the method, and the corresponding significant level changed in the figure. 

DISCUSSION

25. Line 352: To be fair, your results do not show that fuel composition influences fire temperature because you conducted no burn; you’re saying there are implications for fire temperature based on understanding from other research. Don’t overstate your actual results; distinguish those better from suggested implications.

Changed to “Our results reveal that fuel bed composition is a primary factor that influences the environment where fire occurs [26].”

26. Line 360:Change “compositions” to “composition”.

Changed.

27. Line 351-361, 365:There seems to be a lot of emphasis on spread /intensity implications, but not as much about duration. There could be more discussion of that here as you measured fuel bulk density and depth (lines 141-142) as well as associated tree sources.  The duff qualities could be different and that could affect fire outcomes in a number of ways beyond what is discussed. See, for example, Carpenter et al 2021 Ecosystems 24:1059-1074. 1007/s10021-020-00568-7

There were no significant differences in fuel depth and bulk density across aspects and slope positions. Any discussion of effects of ectomycorrhizal association, if this is what you refer to, in its effects on increased fine-root biomass may be superfluous, especially since we have no data to back-up any claims.

28. Line 376, 381:How much of this expected difference in severity would results from simple sun exposure affecting fuel moisture and not differences in fuels. Aspect differences in fuel moisture are enhanced in the leaf-off season when sun angles are low.

This, in fact, is one of the direct causes. The greatest exposure, in both duration and directness, to the sun on south-facing slopes reduces fuel and soil moisture (a source of fuel moisture), and the increased oak composition of the 1-hour and 10-hour fuels (burns hotter than maple; Dong and Williams 2023  https://doi.org/10.3390/fire6080312)

29. Line 391:It seems to be a bit of an overstatement to declare that the simple presence of more oak spp. litter affects fire spread. In drought, when these landscapes burn, litter is normally not particularly limiting—that is, you’ve not shown that fuel is limited, and you haven’t made any direct burn observations. In this entire paragraph, there’s lots of inference. Instead, focus on what this paper actually contributes. You’re not demonstrating fire spread potential; you’re just inferring it from others’ work and your plot data that shows there are some differences in fuel bed properties with certain implications. Avoid overstatement with more careful wording.

We appreciate the recognition of overstating, and have become more specific in our language:

“the percentage of oak fuel is significantly higher on west slopes compared to east slopes, suggesting that west slopes could have a greater potential for higher rates of spread as the result of higher fire intensity and pre-heating of fuels due to the higher flammability of oak fuel compared to other species”

30. Line 394-397:If litter fuel drives fire spread, and it isn’t limiting at this time of year, there is arguably not a “higher level of potential fire”. You might move past this by focusing on potential fire intensity rather than spread potential. That is, what do mean by “potential fire”? The implications of fire on regen or mortality aren’t from just any fire, but only certain types of fire when fuel or fuel moisture may not be limiting.

Changed to “Our results suggest that the upper slope positions have the potential to produce the higher level of fire intensity owing to the high level of oak fuel load and woody plant coverage”.

CONCLUSIONS

31. Line 423: Rewrite: This topic sentence’s emphasis on species distributions is not the focus of this paper and it has been understood for decades. Instead, focus on this paper’s better contribution to fuel patterns and its implications for fire. Fuel is lost even more in sentence two that suggests that the documented species patterns can have consequences for fuel and fire, but fuel has been directly described, not just inferred.

Changed to “This study demonstrated that topographic variables play an important role in species-level fuel bed composition and variation. The different fuel composition is a primary factor that influences the environment where fire occurs, explaining the differences in fire intensities among landscape positions from a fire environment perspective.”

32. Line 428:As mentioned above, this expectation for more fire on south aspects also follows from our understanding of fuel moisture, not just fuel type.

We added other factors: “….. warmer air temperatures, lower relative humidity, and drier fuels”

33. Line 429a:What does “lowest fire” mean? Intensity? Frequency?

Changed to “Among different landscape locations, the south-and west-facing aspects tend to have a higher proportion of oak litter fuels, especially on upper slope positions.  In contrast, the north- and east-facing aspect is expected to have a lower proportion of oak litter fuels.”

34. Line 429b, 430: Wording. Composition “is”, not “are”. Also “data show” not “data shows”.

Changed.

35. Line 429, 430: Nuance the phasing better. Your patterns have implications for subsequent fire intensity. You haven’t demonstrated that because you did not conduct burn observations. This is inferred.

Changed to “Among different landscape locations, the south-and west-facing aspects tend to have a higher proportion of oak litter fuels, especially on upper slope positions.  In contrast, the north- and east-facing aspect is expected to have a lower proportion of oak litter fuels.”

36. Line 431:Confusing statement. Isn’t overstory tree species composition a forest attribute difference? The overstory attributes cause the differences in understory fuel examined.

Changed to “Our data show that there is no distinction between the overstory and understory attributes among aspects and slope positions when not dividing species; however, when divided by oak and maple species, the major differences are found in oak species attributes, and corresponding fuel compositions.”

REFERENCES

37. Line 481, Line 524:Fix cap letters to standardize formatting!

Changed.

38. Line 469, 486, 508, 532, etc.: Fix Use Of Title Caps When elsewhere they are not caps. Standardize formatting!

Changed.

Reviewer 2 Report

Thank you for the chance to read this paper, which sounds like it took a great deal of work. Overall it is well written and interesting, but in my opinion some changes are needed to make the statistics and their interpretation clearer. See below.

Line 22 and 435: “A better prediction” doesn’t seem like the right wording here because the paper is not producing a prediction model for fire behaviour. Some more like “our finding may feed into fire behaviour modelling research” would be more appropriate.

Line 55: The authors should state this applies in the northern hemisphere.

Line 101 to 104: Could you state why some points are missing from the grid work pattern in Figure 1? Were these the inaccessible ones?

Line 167 to 169: Please add a couple of lines about the models. Were the single variable models, simple linear models, GAMs etc?

Line 176, Table 1: Probably would be better if the caption says “Summary statistics” rather than just “observed values”

Line 191/Figure 3: Change “S_position” label in the plot to “Slope position”, or state what it is in the caption.

Figures 3,4,5 and 9: Is the significant difference in slope position displayed in the plots? Are all aspects grouped when testing the significant difference between slope positions, then separated for the plots? Is there a way to highlight which are significantly different, eg by bold or a highlight or a box? Perhaps a table would be better.

 Also, the a, ab etc indication of significant differences for aspect is unclear or takes a while to understand. E.g. is a significantly different to ab? Could you instead highlight only those that are significantly different, e.g. by using the aspect names instead of a, ab etc? This would make it easier to read the text in the paragraph and then see what is happening in the plots. Also, what is the black line on which a, ab etc sit?

Line 226 to 228: Were differences between slope positions by aspect tested? I just wanted to confirm.

Line 270 to 273: Is the interpretation here that maples are shorter and have more understory trees?

Line 299: State the significance level used.

Line 305-306: This sentence is unclear, why does it start with “Besides”?

Line 308-311: Could this sentence be simplified to make the meaning clearer? I don't understand it as is.

Figure 10: Define the variables in the caption – Uba, Utress etc.

Line 316 to 323: Need to explain what the regression analyses are showing and why they were done. Currently, this paragraph only states correlations, but Figures 11 and 12 show R2. It's not clear why these were done, and the results aren’t mentioned in the text. Add some more explanation about these regressions.

Discussion. I would like it made clearer how the results sit in context to the three objectives from the introduction (Lines 84 to 92). I believe this can be garnered from the discussion as is, but it would be better to make this more explicit. Did the results answer each of those questions/objectives?

Line 352 to 353: The results from this study don’t show anything about temperature, only fire environment.

Line 360: Typo – “concluded”

Line 390 to 392: Does this depend on fuel moisture?

Line 403 to 405: What role could fuel moisture play in this?

Good. Only some typos to fix and a couple of sentences could be made clearer.

Author Response

Reviewer 2

1. Line 22 and 435: “A better prediction” doesn’t seem like the right wording here because the paper is not producing a prediction model for fire behaviour. Some more like “our finding may feed into fire behavior modeling research” would be more appropriate.

Changed to “The findings in this study can provide a better understanding of fine-scale fuel bed and vegetation characteristics, which can subsequently feed into fire behavior modeling research in north-central Appalachia.”

2. Line 55: The authors should state this applies in the northern hemisphere.

Changed.

3. Line 101 to 104: Could you state why some points are missing from the grid work pattern in Figure 1? Were these the inaccessible ones?

An explanation regarding missing points on the grid was added in the caption of Fig 1.

4. Line 167 to 169: Please add a couple of lines about the models. Were the single variable models, simple linear models, GAMs etc?

Added in the method, “Simple linear regression analyses were performed to determine the relationship between fuel load and basal area of oak and maple species.”

5. Line 176, Table 1: Probably would be better if the caption says “Summary statistics” rather than just “observed values”

Changed.

6. Line 191/Figure 3: Change “S_position” label in the plot to “Slope position”, or state what it is in the caption.

Stated it in the caption of Fig 3.

7. Figures 3,4,5 and 9: Is the significant difference in slope position displayed in the plots? Are all aspects grouped when testing the significant difference between slope positions, then separated for the plots? Is there a way to highlight which are significantly different, eg by bold or a highlight or a box? Perhaps a table would be better.

Added explanation, “Surface fuels were analyzed by slope positions with all aspects grouped.”

Changed in Fig. 4.5.9, using dashed lines to display the differences in slope positions, corresponding text added in the caption.

8. Also, the a, ab etc indication of significant differences for aspect is unclear or takes a while to understand. E.g. is a significantly different to ab? Could you instead highlight only those that are significantly different, e.g. by using the aspect names instead of a, ab etc? This would make it easier to read the text in the paragraph and then see what is happening in the plots. Also, what is the black line on which a, ab etc sit?

While the double letter (“ab”, “bc”) may be confusing to some, it is an accepted designation to show what data is significantly different and what ones have similarities, especially when multiple data groups are compared. we feel it is important to make distinctions. The black lines are simply to help indicate over which data group the letter refers to.

8. Line 226 to 228: Were differences between slope positions by aspect tested? I just wanted to confirm.

It tested but did not show differences between slope positions by aspect, the corresponding statements added into the results, for example, “Both oak and maple litter fuel load and litter fuel as a percentage of total 1-hour and 10-hour fuel load did not show differences between slope positions by aspect.”

9. Line 270 to 273: Is the interpretation here that maples are shorter and have more understory trees?

Added interpretation in results, “In other words, maple species are shorter and occupy more understory with more under-story trees, higher understory DBH and basal area compared to oak; oaks are taller and occupy more overstory with higher overstory trees, higher overstory basal area and DBH compared to maple.”

10. Line 299: State the significance level used.

The significance level is shown in the caption of Fig 10.

11. Line 305-306: This sentence is unclear, why does it start with “Besides”?

Deleted “Besides”.

12. Line 308-311: Could this sentence be simplified to make the meaning clearer? I don't understand it as is.

Changed to “When considering ground layer, woody plant coverage and max height are significantly positively correlated to overstory DBH and negatively correlated to overstory trees /ha. Considering the correlation between the ground layer and understory, the woody plant coverage and max height are negatively correlated to understory DBH and positively correlated to understory trees /ha.”

13. Figure 10: Define the variables in the caption – Uba, Utress etc.

Caption added in Fig 10.

14. Line 316 to 323: Need to explain what the regression analyses are showing and why they were done. Currently, this paragraph only states correlations, but Figures 11 and 12 show R2. It's not clear why these were done, and the results aren’t mentioned in the text. Add some more explanation about these regressions.

Explanation added, “Across all overstory variables for both oak and maple species, basal area was the most significant factor in predicting the fuel load with a correlation coefficient of 0.725 and 0.540, respectively. A simple linear regression between basal area and fuel load was created and produced an R2 of 0.725 and 0.540 for oak and maple, which means the over-story oak basal area solely can explain 72% of the variance of oak fuel load, and overstory maple basal area solely can explain 54% of the variance of maple fuel load (Figure 11, 12).”

15. I would like it made clearer how the results sit in context to the three objectives from the introduction (Lines 84 to 92). I believe this can be garnered from the discussion as is, but it would be better to make this more explicit. Did the results answer each of those questions/objectives?

More explanation was added to the third paragraph of the discussion, the first three paragraphs of the discussion can answer the three research questions.

16. Line 352 to 353: The results from this study don’t show anything about temperature, only fire environment.

Changed to “Our results reveal that fuel bed composition is a primary factor that influences the fire environment”. Deleted the statement regarding fire temperature.

17. Line 360: Typo – “concluded”

Changed.

18. Line 390 to 392: Does this depend on fuel moisture?

Added explanation, “, the percentage of oak fuel is significantly higher on west slopes compared to east slopes, suggesting that west slopes could have a greater potential for higher rates of spread as the result of higher fire intensity and pre-heating of fuels due to the higher flammability of oak fuel compared to other species”.

19. Line 403 to 405: What role could fuel moisture play in this?

Changed to “Thus, from the perspective of fuel composition, a higher fire intensity should be expected on the upper slope positions among all aspects, especially the south-facing aspects; in contrast, the lowest fire intensity would likely occur on the lower position, especially on the north-facing aspects. This would be exacerbated with drier fuel conditions on upper slopes compared to lower slope positions”.  (Schwemlein and Williams 2007)

Reviewer 3 Report

This work is important for understanding the fire environment, especially in the maple and oak forests in the eastern U.S.. The in-site observation dataset was valuable for studying the topographic and forest influence on fuel attributes. This study evaluates the effects of fine-scale topography and forest attributes on variation within the fire environment by collecting data from 94 plots within an oak-hickory forest across different aspects and slope positions. The authors found that the topographic variables play an important role in variations within the fire environment, and south-facing slopes and upper slope positions with the highest oak fuel load are prone to experience the most intense fires. Overall, this is an interesting study, and the paper is generally well written and organized. 

Specific comments:

(1) Line 17-22: These statements lack clarity and credibility. Though the authors state that “the oak fuel load was highest…” and “Basal area was the most significant factor…” it does not elaborate on how this relationship was determined for the fire environment.

(2) Line 45-47: I think this sentence needs to be reorganized to emphasize the “specific relationship” mentioned above but not “by remote sensing or LiDAR-based data.”

(3) Line 85-92: The objective of this study should be more concise and targeted.

(4) Line 329-331: Maybe the title of Table 5 should describe maple fuel load, maple fuel percent, and maple attributes rather than oak?

(5) Line 416-419: This conclusion is not so convincing.

(6) Line 420-422: It seems inappropriate that Table 6 emerged in the Discussion section.

(7) There are a total of 12 figures in the main text. I suggest relocating some of them from the main text to the supplementary materials, especially those that do not directly support the main conclusion of the study.

suggestions:

add the plot photos to show the fuels in maple and oak forests.

for further study, the topographic position index (TPI) may be a good indicator to classify the topographic position which is also more related to the training in identifying topographic features for firefighters.

Author Response

Reviewer 3

1. Line 17-22: These statements lack clarity and credibility. Though the authors state that “the oak fuel load was highest…” and “Basal area was the most significant factor…” it does not elaborate on how this relationship was determined for the fire environment.

Changed to “The results reveal that fuel bed composition changed across aspects and slope position, and it is a primary factor that influences the environment where fire occurs. Specifically, the oak fuel load was highest on south-facing slopes and upper slope positions while maple fuel loads were similar across all aspects and slope positions. Oak and maple basal areas were the most significant factors in predicting the oak and maple fuel load, respectively.”

2. Line 45-47: I think this sentence needs to be reorganized to emphasize the “specific relationship” mentioned above but not “by remote sensing or LiDAR-based data.”

Changed to “Some studies used remote sensing or LiDAR-based data to quantify forest attributes, monitor forest inventory, and subsequently simulate fire environment and fire risk based on forest attributes and forest fuel [13,14].”

3. Line 85-92: The objective of this study should be more concise and targeted.

Changed, “The objective of this study was to 1) determine how forest attribute variables varied across different aspects (N, E, S, W) and slope positions (upper, middle, lower), 2) assess the relationship between forest attributes, including forest overstory (canopy closure, average DBH, forest density, basal area), understory (average DBH, tree density, basal area), ground layer (herb coverage, woody plant coverage, and maximum shrub height), and fuel conditions (fuel load, fuel composition), and 3) determine how topographic factors affect the fuel composition at the single-species level (oak and maple).  ”

4. Line 329-331: Maybe the title of Table 5 should describe maple fuel load, maple fuel percent, and maple attributes rather than oak?

Changed.

5. Line 416-419: This conclusion is not so convincing.

Removed and changed, “According to the PCA, however, maple species tend to occupy understory with more understory trees, higher understory DBH and basal area compared to oak, suggesting that maple have the potential to overtake oak species. Combined with the results that oak species have a lower number of trees and basal area on east- and north-facing aspects, the process of shifting from oak to mesophytic species tends to occur on the east- and north-facing aspects first.”

6. Line 420-422: It seems inappropriate that Table 6 emerged in the Discussion section.

Changed to results, and corresponding text added in the results.

7. There are a total of 12 figures in the main text. I suggest relocating some of them from the main text to the supplementary materials, especially those that do not directly support the main conclusion of the study.

We believe that all figures are critical to supporting our conclusions, and each are discussed in the text.

suggestions:

8. add the plot photos to show the fuels in maple and oak forests.

photos for maple and oak fuels in the forests are not available but added maple and oak fuel distribution maps in Fig 1.

9. for further study, the topographic position index (TPI) may be a good indicator to classify the topographic position which is also more related to the training in identifying topographic features for firefighters.

TPI was added to the Fig.2 (C)

Round 2

Reviewer 1 Report

This version is greatly improved over the first.

Title/Abstract: Is it “North-Central” or “north-central”? 

Lines 326+, 344. Comment. Figure 10. I like this figure. The added discussion in the text in section 3.3 is quite useful.  There could be more explicit mentions of what Fig. 10 tells us again in the Discussion prior to the heavy emphasis on topography that (over?) dominates that section. This is just a comment.   

Line 391: The reference here [26] is to an earlier related study, not this study, and there should be more distinction about whose results are being discussed. Of course, reader want to make connections to prior published works, but it is confusing when the results of this study are blended with the authors’ prior research this way. You might more explicitly compare results to what is known from that prior study.

Line 392-395:  Check for consistency and clarity to avoid confusion on the part of readers. This summary statement in the Discussion seems inconsistent with that from the results reported on lines 308-314 and shown in Figure 9B that show aspect differences for oak density and basal area: L392: “Our analysis shows that there is no distinct difference in overstory and understory attributes among aspects and slope positions; however, the major differences are found in the litter fuel compositions, in which the south-facing and west-facing aspects have the higher oak litter fuel than other aspects.”

Line 424-431: Rework. Arguably more important than litter is differences in radiation on north aspects. This could be re-acknowledged here too as it’s the fundamental driver of fuel moisture across aspects.  Line 427: this is confusing. Are you referring to understory fuel ladders on the south or north side? Line 439: you can’t say that increased crown fire risk in Australia relates to crown fire hazard in eastern US deciduous forests as there are so many factors that control crowning, and are you referring to N or S aspects here? There are more local Appalachian references that you might consider for ladder fuels, but the best of these relate to table mountain pine, so their relevance would still need nuance. Line 429: How much of “coverage” is functioning as ladder fuels as you aren’t considering the more flammable from the neutral components or seasonal availability? This paragraph could be tightened up.

Line 484-487: Rewrite the final lines using clearer language: The last few lines (almost) restate what’s been said but the message becomes confusing as you actually reported differences by aspect. Here it says …no distinction “when not dividing species…”  This nuance leaves the reader with a murky understanding of the implications of this work.

Reference [24]. This citation is not complete: Fei S, Steiner KC. Evidence for Increasing Red Maple Abundance in the Eastern United States.  

Reference [33]. This citation is not complete: Iverson LR, Dale ME, Scott CT, Prasad A. A GIS-derived integrated moisture index to predict forest composition and productivity of Ohio forests (U.S.A.).   

Reference [36] This citation is not complete: Kolaks JJ, Cutter BE, Loewenstein EF, Grabner KW, Hartman G, Kabrick JM. Fuel loading in the central hardwoods. Published online 2003.

Reference [38] This citation is not complete: Schwemlein DJ, Williams RA. Effects of landscape position and season of burn on fire temperature in southern Ohio’s mixed oak forests. Published online 2007:8.   Is it this? https://www.fs.usda.gov/research/treesearch/27833

Author Response

Please see attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Thank you for the changes and well done on you work. I would only suggest for the boxplot figure captions, mentioning the long dashed lines only appear where significant differences were detected.

I believe the paper can now be accepted.

Author Response

Please see attachment

Author Response File: Author Response.docx