Understory Vegetation Responses to 15 Years of Repeated Fuel Reduction Treatments in the Southern Appalachian Mountains, USA
Round 1
Reviewer 1 Report
Understory vegetation responses to 15 years of 3 repeated fuel reduction treatments in the Southern Appalachian Mountains, USA
Journal: Forests
Key messages
In this manuscript, the authors present the results of a 15 year fire and fire-surrogate study in the Southern Appalachian Mountains, particularly focused on “understory” vegetation communities before and after a suite of treatments. They highlight the success of two treatment types (burning and burning + mechanical tree removal) for restoring desired understory species but a slow response by herbaceous vegetation.
Overall, the manuscript presents some strong findings as part of a larger Fire and Fire Surrogate study project. The research design, methods and results are thorough and robust, although could be clarified according to the specific comments below. The conclusions are supported by the results, however, the contribution of this manuscript to a broader conversation on desired management objectives and the value of different fire and fire surrogate treatments is lost among the extensive results and focus on the vegetation itself. Reframing the introduction to provide context for the desired management objectives and restructuring the discussion to highlight the various success of each treatment type at achieving desired objectives would greatly improve the quality of this manuscript.
MANUSCRIPT COMMENTS
Abstract
Abstract currently contains too many results – pick the most meaningful elements for each treatment type and report those only. Provide a summary statement for each treatment – to what extent was it successful at restoring the desired forest structure?
Line 11-12: Provide some brief context for the primary tree species (e.g., “open oak woodland to closed-canopy mesic forests dominated by…”
Line 20: stems 10-50cm what? (I think this is height rather than diameter at breast height but please specify)
Lines 21, 24, 27, 28: what do you mean “driven by”? It would be helpful to clarify that these are the dominant species
Introduction
Well-written introduction that provides important context for the study. It could be improved by directly linking the desired “open woodland” conditions to management objectives and highlighting why those six vegetation guilds are the key ones of interest. There may be some room for bringing in more restoration ecology literature as the framework for the importance of the research.
Line 47: What are the dominant tree species in the “open woodland” communities described?
Line 62-63: It would be beneficial here to link the species of management concern to the way you have organized the vegetation “guilds” – why are those six guilds most important for helping to restore open woodland communities?
Line 77: Can you expand on the list of fire surrogate treatments? (either here or in the subsequent paragraph)
Line 104 (and throughout): perhaps use “vegetation communities” rather than “guilds”, or explain what constitutes a guild early on
Research Objectives
Line 124: What differentiates a primary from secondary guild?
Line 124: Be consistent in naming – provide scientific name as well as common name for species
Line 132-141: Perhaps group your hypothesis according to treatment type, e.g., B and MB treatments will significantly increase oak cover (white and red), alter pine cover (through increases in yellow pine and decreases in white pine), etc. These are very specific hypothesis and it would be good to see some grounding in the literature for these, even if it they based on the previous results of the FFSS.
Materials and Methods
A revised Figure 1 would greatly improve this section. Also consider using a table to list all of the variables measured rather than a paragraph of information (Lines 179-194)
Location
Line 148: Although you are referring to no past land use in the last 80-120 years, it is very likely that there were Indigenous peoples living on this land prior to European settlement. An acknowledgement of this is very important, particularly given in the West much of the open oak habitat maintained by fire may have been so by Indigenous peoples’ burning practices.
Study Design
Line 167: Choose either “treatment” or “strategies” to describe the management options
Analysis
Line 205: Why divide by 3 for woody stems and only 2 for non-woody species?
Line 219: Introduce (Δ delta) earlier on
Results
A brief reminder before the “Oaks” section saying something along the lines of the “Results will be presented below based on the six primary guilds of interest.”
Within your results, it may be useful to have the first sentence of each paragraph as a summary that outlines the most important findings. Otherwise, you’re repeating the data verbally (in paragraph form), in Table 2, and in the Figures. I think you can rely on the Figures and Tables for all the detail, and use the written results to highlight the most interesting findings.
Line 231: I would change it to say “different from those observed after the M treatment and in the C plots”. Control isn’t a treatment type. (be consistent throughout the results)
Discussion
I would recommend reorganizing the discussion. From a management perspective, the important question is which treatment achieves the desired results. So, the discussion should start with a reminder of the desired results (e.g., “open” woodland with a particular species composition), and then move on to discussing to what extent each of the three treatments (compared to the control and others) helps to achieve those results. There are a lot of measurements (e.g., stem density, absolute cover, IV, Δ abundance) and it is not entirely clear how these measurements link to desired characteristics. It may be worth having a summary figure (or table) that captures the different objectives that each treatment type achieves that could be referenced in the text. Your conclusions summarize this well – use this as the framework for you discussion.
It is also worth including a brief discussion for how topographic variation and the time-since-treatment (that you referred to as limitations early on) may have affected your results. Is there other literature you can draw on to discuss how the treatment types may vary according to these variables?
Oaks
Does the general increase in oak abundance help achieve an “open” woodland vegetation type, especially if they are being recruited into the canopy as suggested by Steiner et al.? There is evidence from oak forests in the Pacific Northwest that over time fire will help maintain open oak savanna conditions whereas the lack of disturbance causes transition to the dense oak woodland (e.g., McDadi and Hebda 2008 Forest Ecology and Management).
Pines
Line 458: Overall increase in pine stem density?
Line 465: Does mechanical + burning necessarily equal a more severe or “large” disturbance? Large is meaningless unless it specifically means area disturbed. I would suggest it is not a higher severity disturbance but rather a higher frequency of disturbance over the timeframe studied.
Mesic Hardwoods, Shrubs, Graminoids, Forbs
No specific comments
Conclusions
Line 587: How do B and MB create the “largest structural changes” compared to the M alone? This may need to be clarified further up in the methods where you describe the treatments.
FIGURES AND TABLES
Table 1
For the MB treatments, please specify which years correspond to which treatments
Table 2
I would suggest that since the (Δ stems/ha) is what you are presenting, that the raw 2001 and 2016 numbers aren’t important especially because they are relatively visible on the figures. For clarity, I would keep your first two columns the same, and then have your subsequent columns be the four treatment types (C, B, M, MB) with the avg Δ stems/ha as the data presented. Otherwise the table is too complex and it is difficult to compare across treatment types. The statistically different lettering could be presented as a superscript to the stems/ha (because this is also visible in the figures).
Lines 265-268: The Table caption could be shortened to just say the “change in abundancy from 2001 to 2016 for the primary and secondary vegetation guilds.”
Figure 1
This figure needs some substantial clarification:
The locator map should be of better quality – just as a topo or aerial photograph
Please label the 3 treatment areas.
Where are the control plots?
Why do the pink and blue lines overlap? If pink is MB than just use pink.
In the second treatment area from the top, why is there an area where the black (B) overlaps with the pink (MB), but not the whole pink? And why is there blue (just M) in the black (B) outline?
The black line on the bottom right isn’t closed
Author Response
Key messages
In this manuscript, the authors present the results of a 15 year fire and fire-surrogate study in the Southern Appalachian Mountains, particularly focused on “understory” vegetation communities before and after a suite of treatments. They highlight the success of two treatment types (burning and burning + mechanical tree removal) for restoring desired understory species but a slow response by herbaceous vegetation.
Overall, the manuscript presents some strong findings as part of a larger Fire and Fire Surrogate study project. The research design, methods and results are thorough and robust, although could be clarified according to the specific comments below. The conclusions are supported by the results, however, the contribution of this manuscript to a broader conversation on desired management objectives and the value of different fire and fire surrogate treatments is lost among the extensive results and focus on the vegetation itself. Reframing the introduction to provide context for the desired management objectives and restructuring the discussion to highlight the various success of each treatment type at achieving desired objectives would greatly improve the quality of this manuscript.
MANUSCRIPT COMMENTS
Abstract
Abstract currently contains too many results – pick the most meaningful elements for each treatment type and report those only. Provide a summary statement for each treatment – to what extent was it successful at restoring the desired forest structure?
L: 11-29
Thank you for this comment. We have reformatted the for clarity and brevity, and have reduced the amount of results reported therein. In its revised form, it focuses on the treatments, key results, and management objectives.
Line 11-12: Provide some brief context for the primary tree species (e.g., “open oak woodland to closed-canopy mesic forests dominated by…”
L: 12-13
Done. Additional details provided, as requested.
Line 20: stems 10-50cm what? (I think this is height rather than diameter at breast height but please specify)
L: 22, 24
Correct. This is total height. We made appropriate edits for clarity.
Lines 21, 24, 27, 28: what do you mean “driven by”? It would be helpful to clarify that these are the dominant species
This sentence was deleted when we revised and shortened the abstract.
Introduction
Well-written introduction that provides important context for the study. It could be improved by directly linking the desired “open woodland” conditions to management objectives and highlighting why those six vegetation guilds are the key ones of interest. There may be some room for bringing in more restoration ecology literature as the framework for the importance of the research.
Line 47: What are the dominant tree species in the “open woodland” communities described?
L: 44-45
We provide a description of the dominant species that would historically have been found in an open woodland
Line 62-63: It would be beneficial here to link the species of management concern to the way you have organized the vegetation “guilds” – why are those six guilds most important for helping to restore open woodland communities?
L: 60-62 and 106-118.
Thank you for this comment. Additional details are added to justify why we are using guilds to measure treatment effects. These guilds are of management interest as they provide insight into future successional patterns that would result in an “open-woodland” forest structure desired by forest managers, and thus are useful benchmarks for management and restoration monitoring. Some guilds represent “undesirable” species (e.g. mesophytic hardwoods), while others are desirable (oaks, herbaceous vegetation, etc.)
Line 77: Can you expand on the list of fire surrogate treatments? (either here or in the subsequent paragraph)
L: 77 and 106. Done.
Line 104 (and throughout): perhaps use “vegetation communities” rather than “guilds”, or explain what constitutes a guild early on
L: 60-62 and 106-118.
Per the previous comment, we explain and justify our use of “guilds” for classification. These guilds represent vegetation groups or species of management interest, and were treated separately because each may respond differently to the different fuel reduction treatments.
Research Objectives
Line 124: What differentiates a primary from secondary guild?
L: 106-118
Our edits more clearly explain what primary and secondary guilds are. Primary guilds generally represent broad groups of species (e.g. oaks), whereas secondary guilds represent more specific groups (e.g. red oaks and white oaks). We felt it was important to analyze secondary guilds, as they provide insight into which species may be driving the change observed for the primary guild.
Line 124: Be consistent in naming – provide scientific name as well as common name for species
Thank you for catching this. We made appropriate edits throughout the manuscript to address this.
Line 132-141: Perhaps group your hypothesis according to treatment type, e.g., B and MB treatments will significantly increase oak cover (white and red), alter pine cover (through increases in yellow pine and decreases in white pine), etc. These are very specific hypothesis and it would be good to see some grounding in the literature for these, even if it they based on the previous results of the FFSS.
While we understand this comment, we were primarily interested in the differences between treatments. We stated the hypotheses this way as it follows the separate analyses, comparing treatments, that were done for each guild. We were concerned that revising our hypotheses at this stage, would require new analyses.
Materials and Methods
A revised Figure 1 would greatly improve this section. If the message you want to carry is about the spread and variability of the data, then standard deviation is the metric to use (Lines 179-194)
Thank you for this comment. We revised figure 1 (the map) for clarity. Specifically we replaced the multi-colored map that did not clearly show the treatment areas with a diagram that shows the general layout of the study design, with an inset diagram that shows the layout of an individual treatment unit.
Yes, we felt that it was important to describe the spread and variability of the data, and that is our justification for reporting SD values in the table.
Location
Line 148: Although you are referring to no past land use in the last 80-120 years, it is very likely that there were Indigenous peoples living on this land prior to European settlement. An acknowledgement of this is very important, particularly given in the West much of the open oak habitat maintained by fire may have been so by Indigenous peoples’ burning practices.
Excellent point. Thank you for bringing this up. On L 134-135 we added a sentence to describe historic land use by indigenous populations.
Study Design
Line 167: Choose either “treatment” or “strategies” to describe the management options
Thank you for catching this. On L 150 we changed “strategies” to “treatments” to match the rest of MS
Analysis
Line 205: Why divide by 3 for woody stems and only 2 for non-woody species?
L: 198-202
We added a sentence to clarify the data available for woody and non-woody species. Stem counts were not collected for non-woody species, thus the importance value calculations for these species were calculated differently.
Line 219: Introduce (Δ delta) earlier on
L: 194-195.
Thank you. We added that sentence to beginning of Analysis section for clarity.
Results
A brief reminder before the “Oaks” section saying something along the lines of the “Results will be presented below based on the six primary guilds of interest.”
L. 225-226.
We added this sentence, as suggested.
Within your results, it may be useful to have the first sentence of each paragraph as a summary that outlines the most important findings. Otherwise, you’re repeating the data verbally (in paragraph form), in Table 2, and in the Figures. I think you can rely on the Figures and Tables for all the detail, and use the written results to highlight the most interesting findings.
Thank you for this comment. Overall we made an effort in this revision to make the results more concise, focusing on the most important findings -- per this suggestion. Additionally, sentences were added to each section that summarize primary and secondary guild findings) However, in some cases, the data showed varied responses by treatment -- nuances which necessitated more detail in our opinion.
See L: 228-230, 255-257, 278-280, 304-306, 312-315, 340-342, 351-352, 363-365, 371-372, 384-386, 391, 403-405
Line 231: I would change it to say “different from those observed after the M treatment and in the C plots”. Control isn’t a treatment type. (be consistent throughout the results)
Thank you for this comment. On L 234, 259, 263, 280, 309, 314, 346, 368, 370, 391, 395, 409 we made appropriate edits for consistency.
Discussion
I would recommend reorganizing the discussion. From a management perspective, the important question is which treatment achieves the desired results. So, the discussion should start with a reminder of the desired results (e.g., “open” woodland with a particular species composition), and then move on to discussing to what extent each of the three treatments (compared to the control and others) helps to achieve those results. There are a lot of measurements (e.g., stem density, absolute cover, IV, Δ abundance) and it is not entirely clear how these measurements link to desired characteristics. It may be worth having a summary figure (or table) that captures the different objectives that each treatment type achieves that could be referenced in the text. Your conclusions summarize this well – use this as the framework for you discussion.
While we understand and respect this reviewer’s thoughts about the discussion, we felt that it was important to keep the discussion organized in the same order as the results. Thus we elected not to make this change. However, we did add a paragraph at the beginning of the discussion that synthesizes our findings from a management perspective.
It is also worth including a brief discussion for how topographic variation and the time-since-treatment (that you referred to as limitations early on) may have affected your results. Is there other literature you can draw on to discuss how the treatment types may vary according to these variables?
Thank you for this comment. In the summary paragraph described above, we provide a brief statement that addresses time-since-treatment and topographic variation. We dive deeper into the results to show that the effects of time since treatment are minimal, and we discuss how fire behavior interacts w/topography.
Oaks
Does the general increase in oak abundance help achieve an “open” woodland vegetation type, especially if they are being recruited into the canopy as suggested by Steiner et al.? There is evidence from oak forests in the Pacific Northwest that over time fire will help maintain open oak savanna conditions whereas the lack of disturbance causes transition to the dense oak woodland (e.g., McDadi and Hebda 2008 Forest Ecology and Management).
L: 460-463. Good point. With regards to developing an open woodland community, the most effective treatment should both enhance oak regeneration and provide recruitment opportunity into larger size classes; therefore, MB is likely the most effective treatment for oak succession over time. So an increase in oak regeneration may not be sufficient to create an open woodland community, but it is a likely pre-requisite.
Pines
Line 458: Overall increase in pine stem density?
L: 488.
This sentence was edited for clarity.
Line 465: Does mechanical + burning necessarily equal a more severe or “large” disturbance? Large is meaningless unless it specifically means area disturbed. I would suggest it is not a higher severity disturbance but rather a higher frequency of disturbance over the timeframe studied.
L: 494-496.
Good point. We deleted the qualifier “large” for clarity.
Mesic Hardwoods, Shrubs, Graminoids, Forbs
No specific comments
Conclusions
Line 587: How do B and MB create the “largest structural changes” compared to the M alone? This may need to be clarified further up in the methods where you describe the treatments.
L: 618-619.
Good point. We changed the sentence to clarify what we meant by “largest” by adding details about the expected sustained/immediate structural changes
FIGURES AND TABLES
Table 1
For the MB treatments, please specify which years correspond to which treatments
L 167. Done.
Table 2
I would suggest that since the (Δ stems/ha) is what you are presenting, that the raw 2001 and 2016 numbers aren’t important especially because they are relatively visible on the figures. For clarity, I would keep your first two columns the same, and then have your subsequent columns be the four treatment types (C, B, M, MB) with the avg Δ stems/ha as the data presented. Otherwise the table is too complex and it is difficult to compare across treatment types. The statistically different lettering could be presented as a superscript to the stems/ha (because this is also visible in the figures).
We made appropriate edits to the table to make it easier to interpret, and we clarified which guild (primary/secondary) was reported in table and figure captions. However, the figures show trends for the primary guilds, while the raw numbers in the table refer to the secondary guilds. Thus we feel that they provide important insight into which species are driving the changes observed in the figures.
Lines 265-268: The Table caption could be shortened to just say the “change in abundancy from 2001 to 2016 for the primary and secondary vegetation guilds.”
Done L: 266-270
Figure 1
This figure needs some substantial clarification:
The locator map should be of better quality – just as a topo or aerial photograph
Please label the 3 treatment areas.
Where are the control plots?
Why do the pink and blue lines overlap? If pink is MB than just use pink.
In the second treatment area from the top, why is there an area where the black (B) overlaps with the pink (MB), but not the whole pink? And why is there blue (just M) in the black (B) outline?
The black line on the bottom right isn’t closed
Done. We replaced the original figure with diagrams that more clearly show the study design.
Reviewer 2 Report
Abstract
Decimals not needed on stems/ha numbers.
Avoid starting sentences with treatment acronyms.
Introduction
Overall, very well written.
Ln56: Change midstory to understory as these species are shrubs.
Ln59: Reword end of sentence to be more direct.
Ln95-100: Most of this info should be moved to Methods section.
Table 1: Move to Methods section.
Ln115-131: Too much background info from previous study, clouds story of this post 15 year study.
Ln132-141: Too wordy. Do not need to repeat “treatments will” for each hypothesis, etc… Make as direct as possible.
Methods
Ln153: Have already provided latin name for mountain laurel. Check other spp. as well.
Figure 1: Need a better locator map. Readers may not be familiar with the western extent of the NC/SC border. Topo map really doesn’t add anything, difficult to see.
Ln167-173: Refer reader to Table 1 at end of this sentence. Again, this info is repeated from initial info in intro. Description of treatments is better suited for Methods section.
Ln174: Delete (ha) and just use throughout manuscript. Explanation of acronym not needed.
Ln191-194: Citation needed.
Ln215-218: Be consistent with either using the latin name or common name of species. In previous text, common names were used.
Results: Text must be reduce. A lot of numerical data are repeated in text and table format.
Ln228: In subsection, include which oaks where observed.
Ln229: Clarify if you are presenting mean ± SD at first occurrence (I had to look at your table to figure out which was used). These values seem extremely high, bringing in to question the validity of statistics. Consider presenting SE rather than the SD as you are testing and reporting means among treatments.
Figure 2 & 3 captions: Need to include that data correspond to oaks in caption.
Table 2 & Results text: Suggest reducing results text. Results become lengthy to read as mean changes are included in both the text and table. Use results text to simply describe trends and refer reader to table. Round stems/ha to whole number.
Ln304: Delete extra tab at beginning of new paragraph.
Ln310: Extra space between mean and SD for B treatment.
Ln344: Delete extra tab at beginning of new paragraph.
Discussion
Ln411: Delete extra space after punctuation.
Ln434: Delete extra tab at beginning of new paragraph.
Ln528: Delete extra tab at beginning of new paragraph.
Author Response
Abstract
Decimals not needed on stems/ha numbers.
Done. L: 21, 22, 24, 26
Avoid starting sentences with treatment acronyms.
Done. L: 20, 22
Introduction
Overall, very well written.
Ln56: Change midstory to understory as these species are shrubs.
L: 44. The development of a shrubby midstory is one consequence of fire suppression. So we kept the word midstory, but added appropriate edits to make the sentence more clear.
Ln59: Reword end of sentence to be more direct.
L: 56. Done.
Ln95-100: Most of this info should be moved to Methods section.
Done. Moved to L: 150-157
Table 1: Move to Methods section.
Done. Moved L: 167-168
Ln115-131: Too much background info from previous study, clouds story of this post 15 year study.
L:103-106 We removed a sentence to reduce the focus on the previous study.
Ln132-141: Too wordy. Do not need to repeat “treatments will” for each hypothesis, etc… Make as direct as possible.
L: 119-126 Done. Repetitive wording was removed for clarity.
Methods
Ln153: Have already provided latin name for mountain laurel. Check other spp. as well.
Done. In the revised manuscript, we mention the common and Latin names the 1st time, then common names only thereafter.
Figure 1: Need a better locator map. Readers may not be familiar with the western extent of the NC/SC border. Topo map really doesn’t add anything, difficult to see.
Done. We edited the map for clarity.
Ln167-173: Refer reader to Table 1 at end of this sentence. Again, this info is repeated from initial info in intro. Description of treatments is better suited for Methods section.
L: 150-157 We deleted this sentence and added the info to Methods.
Ln174: Delete (ha) and just use throughout manuscript. Explanation of acronym not needed.
Done L: 159
Ln191-194: Citation needed.
L: 185 Data were collected according to the FFSS sampling protocol. This info was added in the revised manuscript.
Ln215-218: Be consistent with either using the latin name or common name of species. In previous text, common names were used.
Edits were made throughout the manuscript for consistency.
Results: Text must be reduce. A lot of numerical data are repeated in text and table format.
The results section was revised for brevity and to increase emphasis on the most important trends.
Ln228: In subsection, include which oaks where observed.
L: 230-231. Done.
Ln229: Clarify if you are presenting mean ± SD at first occurrence (I had to look at your table to figure out which was used). These values seem extremely high, bringing in to question the validity of statistics. Consider presenting SE rather than the SD as you are testing and reporting means among treatments.
L: 232. Yes, this is standard deviation. We added “SD” at first occurrence for clarity. We felt it useful to report SD instead of SE, since SD is a better indicator of the spread and variability of the data.
Figure 2 & 3 captions: Need to include that data correspond to oaks in caption.
Edits were made to this caption, and the others, for clarity.
Table 2 & Results text: Suggest reducing results text. Results become lengthy to read as mean changes are included in both the text and table. Use results text to simply describe trends and refer reader to table.
Thank you for this comment. Overall we made an effort in this revision to make the results more concise, focusing on the most important findings -- per this suggestion. Additionally, sentences were added to each section that summarize primary and secondary guild findings) However, in some cases, the data showed varied responses by treatment -- nuances which necessitated more detail in our opinion.
Round stems/ha to whole number. Done
Ln304: Delete extra tab at beginning of new paragraph. Done
Ln310: Extra space between mean and SD for B treatment. Done
Ln344: Delete extra tab at beginning of new paragraph. Done
Discussion
Ln411: Delete extra space after punctuation. Done
Ln434: Delete extra tab at beginning of new paragraph. Done
Ln528: Delete extra tab at beginning of new paragraph. Done
Round 2
Reviewer 1 Report
Thank you for your thoughtful revisions.
