Wildfire and Climate Impacts Tree Hollow Density in a Temperate Australian Forest

Round 1

Reviewer 1 Report

The manuscript is presented interesting. However, there are a few questions and recommendations:

1) There is no clear data on the results obtained in the abstract. Give specific information.

2) Line 183-185. On the basis of what methodology did you determine small, medium and large hollows? Why exactly did you use such dimensions to designate the hollows?

3) Line 188-189. Why was the size evaluated visually? There is a ruler and other tools for this.

4) Table 2 is not readable. Make an correction to the table.

5) I think that within the framework of this study, it is still difficult to say that the climate has affected the tree hollows. It seems to me that the authors should limit themselves to statements about the impact of fires on the hollows and not discuss the topic of climate. In order to discuss how climate change affects hollows, many years of research are needed.

Author Response

Reviewer 1 - Comments and Suggestions for Authors

• R1-C1: There is no clear data on the results obtained in the abstract. Give specific information.

**Authors response: We now include data relating to differences in tree hollow densities per sites between the fire frequency and severity categories in abstract of the here resubmitted manuscript. See lines 21-28:

“A unimodal relationship occurred between medium- and large-sized hollow density and fire frequency and severity (respectively), with hollow densities greatest at intermediate frequencies / severities. Increases of 1.82, 1.43 and 1.17 hollows per site were observed between the 1 (refer-ence) and 2, 3 and > 3 fire frequency categories, and 1.26, 1.75 and 0.75 hollows per site were observed between the low (reference) and moderate, high and very high fire severity categories. Fire severity was also positively associated with basal scar density, with increases of 2.52, 8.15 and 8.47 trees per site between the low (reference) and moderate, high and very high categories.”

• R1-C2: Line 183-185. On the basis of what methodology did you determine small, medium and large hollows? Why exactly did you use such dimensions to designate the hollows?

**Authors response: Hollow dimensions accord with the sizes used by the highly cited reference Gibbon and Lindenmayer (2002; “Hollow formation in eucalypts from temperate forests in southeastern Australia”, Pacific Conservation Biology, 2000, 6, p.218), and many other studies have used these same sized-thresholds in Australian forests. We now acknowledge this in the here resubmitted manuscript by including these references at the end of this sentence (see line 191).

• R1-C3: Line 188-189. Why was the size evaluated visually? There is a ruler and other tools for this.

**Authors response: See comment R2-C3 for a full discussion about this point. Briefly, tree hollows are typically located in the tree canopy, and we were not able to climb trees to actually measure their dimensions. Thus, we used the visual assessment method, which is routinely used in this field of research.

• R1-C4: Table 2 is not readable. Make an correction to the table.

**Authors response: This comment is rather ambiguous. We believe that Table 2 is indeed readable; however, we acknowledge that we did not make any specific references to how the co-efficient estimates for the categorical variables are presented. We have added further information in the Table caption to further describe this in the here resubmitted manuscript. See lines 300 – 301:

“For the categorical variables fire frequency and severity, C.E. are shown as comparisons between a reference group (1 fire frequency, low fire severity) and the other groups (e.g. 2, 3 and 4 fire frequency categories).”

• R1-C5: I think that within the framework of this study, it is still difficult to say that the climate has affected the tree hollows. It seems to me that the authors should limit themselves to statements about the impact of fires on the hollows and not discuss the topic of climate. In order to discuss how climate change affects hollows, many years of research are needed.

**Authors response: We agree that disentangling causal links between climate and tree hollow density is a challenging task. However, we disagree with the reviewer’s premise that the landscape scale associations we present between climate and hollow density are unimportant.  Disentangling such effects often requires multiple lines of evidence from studies conducted at different scales. For example, small-scale experiments can identify probable climatically driven mechanisms driving rates of hollow formation; however, they cannot provide information about whether these processes actually impact hollow densities in nature, especially across larger spatial areas. Conversely, large scale studies like ours can identify landscape scale associations between climatic variables and hollow densities; however, whether these associations are causal requires experimental studies.

In our discussion we highlight the need for future research on climatic vs hollow density effects (e.g. lines 426-434). Further, we are mindful of the correlative nature of our datasets, and thus, discuss our results in relation to this. For example, we describe “relationships” between our climatic variables and hollow densities rather than describing the climatic variables as “predictors” or “drivers” of hollow density themselves. Given the points discussed above, our study provides important results that can act as a “starting point” for other assessments of climate vs hollow interactions. Therefore, we have left climate as an additional focus of our study.

Reviewer 2 Report

General comments

The topic of the manuscript is interesting, and the research aim clear and well focus.

The methodology is well described and scientifically sound, except for one point (please, see below).

The results are clearly presented and useful in the relevant scientific community.

However, there are some points need improvement before acceptance.

Specific comments

Line 105, Please, replace the word “predictions” with the more appropriate “hypotheses”.

Line 118, Please, replace “our study sites” with “the study sites”.

Line 188, Authors write “The size of all hollows was assessed visually by the same observer (Harper et al., 2004)”. However, Harper et al. (2004) suggest (in page 298 and 299) a double-sampling method that allow more efficient inventories of the occurrence of hollows in forests. Please, give a strong explanation to justify your choice.

Line 288, Please, add the area in which the given hollow densities refer to (e.g., per ha).

Conclusions: It is not common to exist references in this section. Please, modify appropriately.

References: Please, check carefully. For example, Harper et al. (2004) is absent in the reference list, and therefore, it should be added.

General comments

The topic of the manuscript is interesting, and the research aim clear and well focus.

The methodology is well described and scientifically sound, except for one point (please, see below).

The results are clearly presented and useful in the relevant scientific community.

However, there are some points need improvement before acceptance.

Specific comments

Line 105, Please, replace the word “predictions” with the more appropriate “hypotheses”.

Line 118, Please, replace “our study sites” with “the study sites”.

Line 188, Authors write “The size of all hollows was assessed visually by the same observer (Harper et al., 2004)”. However, Harper et al. (2004) suggest (in page 298 and 299) a double-sampling method that allow more efficient inventories of the occurrence of hollows in forests. Please, give a strong explanation to justify your choice.

Line 288, Please, add the area in which the given hollow densities refer to (e.g., per ha).

Conclusions: It is not common to exist references in this section. Please, modify appropriately.

References: Please, check carefully. For example, Harper et al. (2004) is absent in the reference list, and therefore, it should be added.

Author Response

Reviewer 2 - Comments and Suggestions for Authors

• R2-C1: Line 105, Please, replace the word “predictions” with the more appropriate “hypotheses”.

**Authors response: Suggested change added to the resubmitted manuscript (see line 109).

• R2-C2: Line 118, Please, replace “our study sites” with “the study sites”.

**Authors response: Suggested change added to the resubmitted manuscript (see line 122).

• R2-C3: Line 188, Authors write “The size of all hollows was assessed visually by the same observer (Harper et al., 2004)”. However, Harper et al. (2004) suggest (in page 298 and 299) a double-sampling method that allow more efficient inventories of the occurrence of hollows in forests. Please, give a strong explanation to justify your choice.

**Authors response: The double-sampling approach suggested by Harper et al (2004) uses periodic “tree climbing” surveys to correct bias’s in hollow detection made using visual assessment methods. Although Harper et al (2004) suggest the double-sampling approach, their data also shows strong positive associations between the visual and tree climbing surveys (which theoretically detected all hollows without error) they used to assess hollow frequency (i.e. an R of 0.70 to 0.83 for experienced observers). Further, a low level of between-observer bias was present in their visual estimate surveys of hollow frequency, when only experienced observers were considered. For example, see section 3.2.2. from Harper et al (2004):

“Correlations between the number of hollows detected through climbing and ground-based surveys were positive. For eight of the nine ground-based surveys, the correlation ranged from 0.70 to 0.83, with a single lower value of 0.57 (Table 3). Five of the nine ground surveys differed significantly from a GLMM formulated from pooled data (Table 4).”

Given the above, we were justified in using visual assessments of tree hollow frequency in our study. This is because our survey was made by the same experienced observer (CEG), which Harper et al (2004) showed should detect most hollows. Further, site-specific hollow-bearing tree and hollow frequency were highly correlated at our sites (R > 0.9). Harper et al (2004) showed stronger correlations between their visual and climbing surveys for hollow-bearing trees than for hollow frequency. Therefore, the results from our study should be more comparable to the later than former from Harper et al (2004).

We initially preferred the double-sampling approach suggested by Harper et al (2004); however, logistical constraints on the field surveyed negated its use (e.g. time constraints, institutional constraints related to Work Health and Safety). We have altered the text in the here resubmitted manuscript to briefly include some of the points discussed above:

See lines 190-193: “The size of all hollows was assessed visually by the same experienced observer (CEG; up to 30 min per survey). We initially preferred the double-sampling approach suggested by Harper [29]; however, logistical constraints on the field surveyed negated its use here.”

• R2-C4: Line 288, Please, add the area in which the given hollow densities refer to (e.g., per ha).

**Authors response: Suggested change added to the resubmitted manuscript, both in Fig. 1 and Fig. 2.

• R2-C5: Conclusions: It is not common to exist references in this section. Please, modify appropriately.

**Authors response: We disagree. In our view, the conclusions section should summarise the primary results in relation to broad scale literature sources, which is what we have done in our manuscript. Either way, I feel this is a minimal stylistic issue. Given this, we have left the references in the conclusions section of the here resubmitted manuscript. 

• R2-C6: References: Please, check carefully. For example, Harper et al. (2004) is absent in the reference list, and therefore, it should be added.

**Authors response: Thank you for this observation. We now include Hunter et al (2004) to the reference list in the here resubmitted manuscript and have ensured that all other references are also included (see lines 519-520).