Spatiotemporal Dynamics and Factors Driving the Distributions of Pine Wilt Disease-Damaged Forests in China

Round 1

Reviewer 1 Report

Comments to the authors.

Overview of the research.

In this MS, the authors studied the spatio-temporal changes of damaged forest areas caused by Pine Wilt Disease. Using a species distribution model with climate, socio-economic and spatial inputs, the authors analyzed the annual change of damages area from 1982 to 2020. The authors found that PWD occurrences in the previous year were the most important variables to explain a current year occurrence, but DEM and Bio14 (precipitation in the driest month) were also important in the years with a significant increase PWD occurrence. The authors concluded that their findings could help identify potential high-risk areas for better forest management and pest control.

General comments:

Overall, I found this paper is worth publishing. The aim of the study is well described and justified, sound data preparation and analysis, clear and well-explained results, and reasonable discussion and conclusion. I like their approach using a species distribution model to analyze annual changes of a forest disease. The result, the current year occurrences were strongly associated with the previous year occurrences of the disease, is not surprising, but their approach is novel in the field of species distribution model. There are, however, some places I can not follow their discussion, and some methodological details are missing. I address three major issues here that I hope the authors can handle without much effort.

• Spatial input data for damages forest area

The authors used PWD damaged forest as a spatial input but did not explain how those data are obtained and processed. Since this parameter is the most important variable for the model, please explain it in detail. Judging from the figures, the authors create a distribution map using district occurrence data, but then the spatial correlation can be measured only by the occurrence of the same district in two consecutive years but not in surrounding districts. How could the model implement the occurrence of PWD in the surrounding district (or cells)?

• Maxent settings

Maxent software offers various options, but the authors did not describe them. Please specify those details. If the authors used a default setting, it would be nice to show a response curve for each parameter since the automatic function could choose a hinge function that can result in over-fitting of each parameter.

• References

I feel some references are cited not that precisely. For example, in L27 and L41, the authors cited Matsuhashi (2020) for "most severe forest diseases worldwide (L27)" and "cold temperatures retard its spread and delay its life cycle(L41)" and "so pine species affected by water deficit are highly susceptible to PWD outbreaks (L46)". It may be true that Matsuhashi et al. mentioned such things in their paper, but their paper is about using an SDM to predict the occurrence of PWD in the northern part of Japan, not a global analysis nor life cycle of nematodes nor water deficit (L46). Matsuhashi clearly mentioned that they could not find any evidence for that assertion and even concluded: "annual precipitation in Japan are therefore not considered a critical factor in PWD risks (P9)". Similarly, In L30, the authors refer to Robinet (2009) and Ikegami (2019), but their studies did not report the spread of PWD in each country. For the occurrence of PWD in Portugal, for example, Mota (1999) is a better publication to cite. In L43, the authors referred paper from Ma (2016), but they studied a vector of PWD under cold temperature, not the eco-physiology of PWD development in a tree. Since Ma suggested -10C as isotherm for the northern limitation as a distribution of vector occurrences, the authors should clarify this if they want to keep this citation.

Minor comments

In L48, the authors said, "climatic factors associated with its distributions are changing", but the distribution area of PWD in China is mainly within the prediction by Ikegami (2019), so it is not possible to say the climate factors are not changing. Instead, the PWD in China is just filling the climate niches, I guess.

From L296- Did you use all occurrence points for model building? Judging from figure 5, the authors used occurrence data from each year but not clearly describe it. Please explain it in the method section.

L331-333 it would be nice to have references to figure 7 here, like "412 county districts (Figure 7)" and "Liaoning, and Chongqing (Figure 7)"

Line-by-line comments.

L54 "climate is measured"> I think you want to say "climate is characterized"?

L80 [4,15] > I think Jaime (2019) is about climatic suitability for host and bark beetle, not PWD. You should replace "PWD" in L79 with forest pest?

L154-156 "These differed among years as a result of short-term variations in the annual environmental variables"> did you repeat the correlation test for the dataset from each year? If so, please clarify it.

L169 "low-low or high-high"> what do you mean by low and high? Risk? Occurrence? Density?

L181 > "statistical significant relationships"> I guess, Maxent does not measure statistical significance. Remove "significant".

L183 "these data"> which data? Outputs?

L295 "that the environmental variables were not selected in the MaxEnt model for that year" > selected by Maxent I thought the authors said based on Pearson's correlation coefficients test in L149-152?

L355-356, how did you process the effects of neighbouring cells? (see comments above)

L365 "most PWD studies have been conducted in a specific year, ….  over 1982–2020 [2]." > are you sure you are citing the right one?

L378 "and Bio11 (mean temperature in the coldest quarter)">  duplicated.

Table 1. > NTL should be spelt out.

Figure 3 > maps within a box are not particularly informative as no additional data are from those areas. Remove them.

Mota, M. M., Braasch, H., Bravo, M. A., Penas, A. C., Burgermeister, W., Metge, K., & Sousa, E. (1999). First report of Bursaphelenchus xylophilus in Portugal and in Europe. Nematology, 1(7), 727-734.

Author Response

Thanks for your comments, which have helped us to improve the manuscript substantially. Many inaccurate words and phrases have been carefully rewritten, and the incorrect citations of the manuscript have been changed. We have provided more detailed information about the Maxent settings. We truly hope that the revised manuscript fully explains all your concerns.

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear authors, I read your manuscript “Spatiotemporal dynamics and factors driving the distributions of pine wilt disease-damaged forests in China” with great interest, and I think this is a very interesting study. You introduce the problem, the spread of pine wilt disease in the forests of China, and perform your analysis, model the data to try to understand the changes in the forest. I think you did a great job, and I believe this is going to be of interest to the public that will be interested in learning more about the spread of PWD forests and use the same method in different areas, considering this is becoming a great challenge for countries that need to deal with invasion of plant diseases, exotic pests, and so on.

I have a few minor comments and suggestions for you, listed below:

Line 50-57: I did not understand what you mean with this explanation because climate is usually the average of a long period of time, however climate data are used yearly, monthly, daily, depending on the study. Are you saying instead of using one single value which would be the average for ~ 30 years, you instead are using the monthly observations for each year?

Line 285-286: But are these changes positive or negative precipitation anomalies?

Line 327: You should reference this sentence to Figure 7.

In Lines 387-395: Do you think that maybe the human activities are being underestimated because of the data you are using? And if it’s not climate or human activities, what explains that huge positive trend from 2008-2018? You think that PWD accelerated distribution from 2008-2018 might be related to biotic interactions?  It would be nice to see citations for this behavior if you have some to cite.

I suggest you add to your figure 2 which factors in your model were most important for each of the outbreaks you show there. To give more information to the readers of what was going on during those periods.

Author Response

We greatly appreciate the comments. We have carefully rechecked and revised the whole manuscript according to your suggestion, and the corresponding changes are marked in the revised manuscript.

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

no major issues

Author Response

Thank you very much for your affirmative comments. We have corrected minor typing mistakes and grammatical errors in our manuscript.