Limited Effects of Long-Term Repeated Season and Interval of Prescribed Burning on Understory Vegetation Compositional Trajectories and Indicator Species in Ponderosa Pine Forests of Northeastern Oregon, USA
Round 1
Reviewer 1 Report
This manuscript described responses of understory plant communities to low intensity prescribed fire seasonality and burn interval. Overall, I found the article well-written overall, the methods and study design well-conceived, and the conclusions reasonable. However, I do have concerns with some methodological choices, and some aspects of the discussion/conclusions. I detail my comments by section below.
Title
I suggest making the title more informative. It seems the authors could make it more of an informative sentence given the (fairly) clear results. Also, “long-term” seems subjective.
Methods
L210: Can the authors justify removing rare species (species present in <5% of plots)? I understand this is a common practice, but rare species can provide very interesting signals, there are transformations that help deal with rare species dominating community analysis signals (see comment below), and rare species seem relevant to this study. Without justification, removing rare species seems arbitrary and does not square with the introduction/objectives as written (i.e., there is no indication in the introduction or objectives that only common species are of interest/importance).
L213-215: The authors need to back this claim up with citations. The Hellinger transformation is typically used prior to community analyses that use Euclidean distances, and it controls for rare species—so I do not understand the authors’ contention here.
Discussion
L393-394: Per Figure 1 and the results regarding initial (and ending) similarities between fall treatments and spring/control treatments, one major caveat (or perhaps very interesting result?) is that spatial autocorrelation may be driving community similarity and trajectory patterns more than fire regime. I realize Figure 1 only shows one example of a stand study design, but the placement of treatments in this example leads me to believe that space may supersede other factors. I see the treatment divisions were randomly assigned (L177-178), but this does not assuage my spatial autocorrelation concerns. The authors need to tackle this potentially major limitation of the study in the discussion and test for spatial autocorrelation in their results—even if the results end up in the Appendices. I would also suggest providing a map of all stand/treatment arrangements in Figure 1.
Conclusion
L427-428: This is a great sentence and (from my perspective) the most important conclusion. I typically do not like commenting on writing particulars, but seeing as this was the core implication, I suggest making this the topic sentence of this paragraph.
L431-438: The authors either need to provide citations for all these statements or cut these lines.
L438-442: These lines seem quite extraneous—like a preview of a future paper? I suggest cutting them.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Dear Authors,
I found your study carefully designed, thoroughly presented, and your inferences mostly well supported by data. I was surprised to note such small changes between treatments. You report surprising results that prescribed burning treatments had similar effects on understory species. You correctly note that this suggests the season of burning will have small effects on management outcomes.
While I think that you have strong reasons to make your main claim in this article, I have one concern: could the weak differences between treatments reflect random differences, rather than responses induced by your treatments? I am not asking this to necessarily change any section of your article, but I would like to at least make sure that my concerns are trivial or negligible. I am worried your results might reflect some sort of artifact because:
- Montia prefoliata's indicator value increases every 5 years in all treatments, including the control. This doesn't make much sense, as we expect Montia's abundance to increase in response to fire.
- Table 2 seem to shows that the two Fall treatments have angles of community trajectories that are similar in 2004, 2007 and 2009. However, the Fall 15YR treatment only gets burned in 98 and 13/14. This might explain why some indicator values (Bromus, Ceanothus, Cirsium, Elymus) are large in both fall treatments, regardless of burn years.
- Figure 3 seem to suggest that the trajectories of all treatments are similar (they move towards the lower left corner of the multivariate space, then move towards greater values of PCoA2), and that there are strong differences in starting conditions.
It is plausible that 5 vs. 15 burning interval might not do as much as fall vs. spring treatment. However, I find it confusing that the two fall treatments seem to fluctuate in unison, and I think this requires an explanation in the manuscript text (or, if I have misunderstood something, of course you won't need to change the text of the article). The main reason to ask for this clarification is that if these results are arising from chance, future studies should not cite this article to give its results a causal explanation.
I also have other more minor points of concern. These are:
- The introduction is a bit too too long. I think that especially the last paragraph could be cut.
- Table 2-3: why do you talk about "triplets" in Table 2, but then about "two-year segments" in Table 3? Is there a reason for using triples and then year pairs?
- Field Sampling: Does "six plots in each unburned control unit" mean that control units alone had twice as many plots? If so, why? What do you mean "systematic sampling"?
Minor comments ----------
37-40: unclear wording - what do you refer to? Why interval and seasonality would be an operational burden? What do you mean by "Alter the importance of seasonality"? This last sentence of the abstract consistently confounded me (I have read it three times), so I strongly suggest to clarify it.
55: why talking about "Driveway sites" instead of "eastern sites"? I propose to only talk about "Eastern sites". This goes to the point, while "driveway" is unnecessary, and unintuitive.
187: Which of the stands was burned in 2012?
211-213: this sentence is not needed - I suggest removing it.
237-239: citation?
249-250: I see "no evidence" rather than "little evidence" that community composition differed between burn intervals. Am I correct? The lowest P-value is 0.39.
264-266: Only for the Fall burns it seems?
298-299: Why do you talk about directionality here?
423: "There are increasing calls to increase": repetition.
429-436: I do not understand this jump to talking about homogeneity in mechanical treatments. I do not think you provide evidence, in the article, to talk about homogeneous thinning versus spatially variable tree retention. I suggest therefore removing this part.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors responded adequately to my comments.
