Modelling Current and Future Potential Habitats for Plantations of Eucalyptus grandis Hill ex Maiden and E. dunnii Maiden in Uruguay

Round 1

Reviewer 1 Report

Line 28 to 31: any ecological or biological reason? Why your correlation exercise shows that? You did not consider the adaptation factor in your study, hence, you should be very careful when stating such a strong comment.  

Fig.1 caption & legend is not appropriate. It does not tell anything useful to the reader. Change pls.

Fig.1, “2, 7, 8, and 9” must be classified as one group, unless you rank them in the model

L80: SDM “full name”

L138: use of 19 bioclimatic is wrong, some are highly correlated with each other. Did you check? Later use stated that yes, so please change your statement here to do not mislead readers.

L 156: .It   add space

Have you had any strategies to reduce the influence of potential biases in the occurrence points of the species you modelled? Refer to https://doi.org/10.1111/ecog.04530 and https://onlinelibrary.wiley.com/doi/full/10.1002/ece3.2332

RCP 85 or 8.5? Apply the same for others pls

In Fig. There are outliers, correct your database so then you do not see them in your plot

The explanation of how you generated your absence data is not clear. This is very important and must be stated.

Any limitation on your study?

Author Response

Please see the attachment:Cover_letter_reviewer_1.docx

Author Response File: Author Response.pdf

Reviewer 2 Report

General Comments

The manuscript entitled “Modelling current and future potential habitat distribution for plantations of Eucalyptus grandis and Eucalyptus dunnii in Uruguay” by Resquín et al. constitutes a valuable contribution on the important issue related to Eucalyptus plantations. The authors employed a species distribution model (SDMs) and environmental variables (climate, edaphic and topographic) to predict the long-term suitability of potential forest areas in Uruguay for two Eucalyptus species under different climate change scenarios. Species showed different sensibility to climatic and soil parameters. Authors use ensemble modeling techniques, and many of the technical aspects of species distribution modeling appear to be well addressed using powerful statistical tools.

However, I suggest some recommendations to authors to improve clarity and confidence on the results:

1. Although I think the authors did a good job in applying some nice existing modelling techniques to assess current and future habitat suitability for two Eucalyptus species, authors should highlight what novel information this manuscript delivers in relation to current and climate change effects on the habitat suitability of Eucalyptus plantation in Uruguay. Why SDMs should be used to answer the main authors’ hypothesis?
2. Being the Eucalyptus introduced species in Uruguay, the distribution of these species represents an “artificial niche” rather than an “ecological niche”. I'm concerned that the presence/absence points don't sample the full range of variables values that occur in the country. Therefore, authors need to justify the methods used for the presence/absence data.
3. Related to the previous point, one of the limitations when applying SDMs is related to the reliability of obtaining models for generalist species (Papes and Gaubert 2007; Mateo et al., 2011). According to eucalyptus species description included in lines 74-77, the site requirements of eucalyptus species are wide enough to achieve good conditions in different regions of Uruguay. Could this issue affect the reliability of the models and the results obtained in this work?. Please, clarify.
4. Other main aspect to be improved is related to the implications for forest management. On my opinion, Discussion should include a “Management Implications” section focused on the how the use of SDMs models may improve the intensive silviculture in Uruguay.
5. The discussion section should also be revised since it repeats many descriptive sentences and only a few paragraphs are truly devoted to put the results and discuss into the context of existing literature.
6. A final aspect to be included in the Discussion section is the reason to include soil and other physiographic variables in the SDMs models. SDMs would be calibrated with information across the full spectrum of conditions spanned by the full ranges of model parameters. The model for Eucalyptus grandis depends almost entirely on soil parameters which may not be a very useful variable when it’s based on a coarse-scale grid. It is not surprising that such a model indicates low sensitivity to climate change. We understand the ambition to make a parsimonious model (though note that this isn’t always a useful ideal), but this should not lead us to exclude the processes we’re interested in from the analysis (i.e. throwing out all parameters that are likely to respond to climate change).

On the other hand, there is another issue that concerns me as well. If the edaphic and topographical variables were considered as constant for future projections, and the model for the prediction of habitat for E. grandis depends on these variables; what predictive capacity can be given to the model for E. grandis in its future projections?. The results showed in table 4 seem to confirm our expectations and the probability of occurrence for E. grandis in all evaluated future scenarios remains unchanged from the present. Please, clarify.

Specific Comments

Abstract

7. Include some statistical information to support your results

Introduction

8. L67-73: This paragraph is not well written.
9. L74-77: Provide a reference.
10. L78-93: The paragraph devoted to introducing SDMs can be improved by doing a reference to the use of these models with non-native species.

Material and methods

11. L122-125: Please improve the details about the national forest inventory of Uruguay.
12. L140-147: Why did you use NCAR-CCSM4 scenarios?

Results

13. Include a sentence doing clearer how the comparison between current and future has been made.

Discussion

14. See comment#4 and 5

Figures

15. Figure 1. Please include a reference (web site) where map information could be consulted.
16. Figures 5 and 6. “RCP 26, 45, 60, 85” must be “RCP 2.6, 4.5, 6.0, 8.5” in the figures and its captions

Tables

17. Table 4. “CW” must be “WD”

Supplementary Material

18. Please include in the supplemental materials a table which identifies all the variables initially considered, which were dropped due to collinearity, and which were utilized in each set of models.
19. L 452. “RCP 26, 45, 60, 85” must be “RCP 2.6, 4.5, 6.0, 8.5”

 

Author Response

Please see the attachment: cover_letter_reviewer_2.docx 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you for addressing my concerns.

Reviewer 2 Report

Thanks for your last modifications. In my opinion significant improvement has been made. The principal suggestions to improve clarity of the manuscript have been addressed and, for example, a management implications section have been included.

On the other hand, although one of the major concerns about the manuscript persists (limitations of E. grandis model), it must be recognized that the authors have made an effort of clarification and transparency and the limitations on this model have been reflected in the last version of the manuscript. Furthermore, the authors propose alternatives to overcome this problem in future models with these species.