Distribution Dynamics of Diplopanax stachyanthus Hand.-Mazz. (Mastixiaceae) and Its Implications in Relict Mastixioid Flora Conservation

Round 1

Reviewer 1 Report (Previous Reviewer 1)

The errors that I pointed out last year have been corrected in the manuscript. There are only some minor formal issues, particularly in the bibliographical references, I have pointed out all the ones I have seen, but it would be advisable for you to check that section and adapt it to the journal's rules.

Comments for author File: Comments.pdf

Author Response

  Based on your suggestion, we have revised the manuscript. Please refer to the reply document for details. We would greatly appreciate any constructive suggestions and comments from the editors and reviewers to further improve our manuscript that strives to meet the strict publishing standards of Forest. If you need any other information, please feel free to contact me at any time.

Author Response File: Author Response.pdf

Reviewer 2 Report (Previous Reviewer 2)

The manuscript, while still using a very repetitive formula:

GBIF+WorldCLIM+MAXENT (or other models based on the entropy of systems).

However, it is quite improved, providing new citations and interesting conclusions about the ecology of D. stachyanthus, which increases the interest of the manuscript. However, there are still some small details or doubts that need to be clarified.

Detailed comments can be found in the attached file.

Comments for author File: Comments.pdf

Author Response

Based on your suggestion, we have revised the manuscript. Please refer to the reply document for details. We would greatly appreciate any constructive suggestions and comments from the editors and reviewers to further improve our manuscript that strives to meet the strict publishing standards of Forest. If you need any other information, please feel free to contact me at any time.

Author Response File: Author Response.pdf

Reviewer 3 Report (Previous Reviewer 3)

Dear authors, 

Well, I see you have improved the manuscript strongly. Some remarks are still necessary to say, however I hope you will easily be through with them.

Figure 2: All visible dots shown are black and green simultaneously. Is the explanation on filtered points necessary?

line 205: Why do you write "Maxent's predicted map" when you remade the modelling with Random Forest (as said in lines 195, 201)? 

Tables 1 & S4: To exclude further misunderstanding, it is worth to name the variable bio2 like "Mean Diurnal Temperature Range (MDTR)". 

line 89: "Cenonzic Northern Hemisphere"??? Do you mean Cenozoic Northern Hemisphere?

line 94: what is "E-O boundary"?

line 171: "we used the average model of the three models to predict ...". Well, maybe it's just a matter of grammar. Whether the following formulation is right: "we averaged climatic variable values calculated by the three selected models to predict ..."?

Author Response

Based on your suggestion, we have revised the manuscript. Please refer to the reply document for details. We would greatly appreciate any constructive suggestions and comments from the editors and reviewers to further improve our manuscript that strives to meet the strict publishing standards of Forest. If you need any other information, please feel free to contact me at any time.

Author Response File: Author Response.pdf

Reviewer 4 Report (Previous Reviewer 4)

Dear authors,

Thank you for your responses to my comments. I mostly accept them, but I still need to see Table S1_1 in Supplementary, not as attached to your response only.  There should be a proper reference to this table in the main text. Besides, titles of the columns in this table should be translated to English.

I suggest also some linguistic editing to your Introduction. Please see these suggestions in the attached file.

Best regards,

Igor Bartish, PhD

Comments for author File: Comments.pdf

I suggested some minor editing to Introduction in the attached pdf file.

Author Response

Based on your suggestion, we have revised the manuscript. Please refer to the reply document for details. We would greatly appreciate any constructive suggestions and comments from the editors and reviewers to further improve our manuscript that strives to meet the strict publishing standards of Forest. If you need any other information, please feel free to contact me at any time.

Author Response File: Author Response.pdf

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

The manuscript is very good and free of methodological errors. Only three major and several minor corrections can be objected to. 

A very common phenomenon occurs in this paper: promising in the title to address items that are then addressed very superficially. I am referring to the end of the title “and its implications to relict mastixioid flora conservation”. However, the deficiencies caused by this fact can be corrected.

The most serious deficiency of this paper is the clear lack of definition of the subject of the study. I am not referring to the species Diplopanax stachyanthus, whose taxonomic location is adequately treated, but to “mastixioid flora”. There is no definition of this concept. What is it? Who proposed it? The citations of the “classical” German authors (Kirchheimer 1957; Mai 1981, 1989) and the establishment of a temporal framework for the mastixioid flora are missing.

In addition to the aspects mentioned above, the living fossils of the mastixioid flora should be described and treated in more detail. It is true that the third paragraph talks about “Known as living fossil plants, relict plants are characterized…” But this is confusing, it can refer to relict plants from any period. The question arises: the authors mean that

Is “relict plants are characterized…” equivalent to “mastixioid plants are characterized…”?

If the answer is yes, please change it. Even with this change the treatment of living mastixioid fossils is very cursory. It would be interesting to know some examples and some numbers on the number of taxa of this peculiar flora.

Recapitulating all the above, the mastixioid flora, referred to in the title, is not well treated either in quality or quantity. To remedy this, I propose to the authors that, after the 2nd paragraph of the 2nd page, there is another paragraph that: defines the mastixioid flora, incorporates the suggested citations, highlights that there are species of this flora that still survive as living fossils. For the study area, it would be convenient to mention some examples and give the number of species that make it up (or an approximate number, or an interval). The ideal would be to attach a list with the scientific names of these relict plants in supplementary material (obviously there may be doubts whether a species really belongs to mastixioid flora, but an approximate list would be very useful).

The references whose introduction is recommended are:

Kirchheimer, F. Die Laubgewiichse der Braunkohlenzeit; Veb Wilhelm Knapp Verlag, Halle (Saale), Germany, 1957; pp. 1–783.

Mai, D. H. Entwicklung und klimatische Differenzierung der Laubwaldflora Mitteleuropas im Tertiär. Flora 1981, 171, 525–582.

Mai, D. H., Development and regional differentiation of the European vegetation during the Tertiary. Plant Syst. Evol. 1989, 162, 79–91.

A second topic where the paper can be improved is that of “implications to (…) conservation”. The authors commit in the title to give importance to plant conservation in the manuscript. But that has little reflection in the Introduction. To correct it, I propose to the authors that after the paragraph that has been proposed previously, they add another one (which would be the 4th on the 2nd page), commenting on examples of threatened plants, as well as the number of these. It would be good if the authors provided the number of species of the mastixioid flora in the study area, evaluated at the international level (IUCN Red List) or national level (Threatened Species List of China's Higher Plants or another more up-to-date catalogue). Let them say how many there are for each level of risk. It would be ideal if this information is captured in a table that is provided as supplementary material, perhaps together with the table of mastixioid species suggested above.

The lack in the conservation area also reaches the species D. stachyanthus itself. Nothing is said about the conservation of this species in the Introduction. That is not acceptable. Only in Discussion, in the 3rd paragraph on page 11, it says that “is currently listed on the IUCN Red List.” Without saying the risk category or the criteria (!) and without including the IUCN sheet as a bibliographic reference (!!). It seems that the authors are not aware that red list conservation status assessments, with their categories and criteria, are a fundamental element in Conservation Biology. The authors opted for a plant conservation approach right from the title. Then they must evaluate conservation status assessments.

For all the above, I propose that the beginning of the 3^rd paragraph on page 11 be omitted, that its content be taken to “Introduction” after talking about the conservation of the entire mastixioid flora. Add a paragraph starting with the beginning of the discussion paragraph:

“D. stachyanthus, an endangered species native to China and Northern Vietnam, is currently listed on the IUCN Red List”, but the authors must continue that the species has been classified as Vulnerable (VU A1c), which means that in 1998 there was an observed, estimated, inferred or suspected population size reduction of ≥50% over the last 10 years, based on a decline in area of occupancy, extent of occurrence and/or quality of habitat (IUCN 2012). In addition, the population has been severely fragmented and there has been a continuing decline of mature individuals (“In China, declines in population numbers have been recorded in the last decade” (1980-90's). The expansion of agriculture (annual) was mentioned as the main threats. & perennial non-timber crops and livestock farming & ranching) and biological resource use (Logging & wood harvesting) (sun, 1998).

If the authors have doubts or disagree with the conclusions of Sun (1998), it is in this paragraph where this interesting debate should be raised.

Suggested references are:

IUCN. IUCN Red List Categories and Criteria: Version 3.1. 2^nd edition. IUCN: Gland, Switzerland and Cambridge, UK, 2012; pp. 1–32. Available in: https://portals.iucn.org/library/sites/library/files/documents/RL-2001-001-2nd.pdf

Sun, W. Diplopanax stachyanthus. The IUCN Red List of Threatened Species 1998, e.T32339A9699334. https://dx.doi.org/10.2305/IUCN.UK.1998.RLTS.T32339A9699334.en.

I would like to make a third and final observation of major importance from the manuscript. It refers to the end of the discussion and the conclusions. This is because section 4.3 refers only to D. stachyanthus. On the other hand, both the title and the conclusions (“To secure the long-term survival of D. stachyanthus and the ancient mastixioid flora,”) talk about the conservation of the mastixioid flora. There is a jump from the species to the floristic group. The problem is that this jump has not been made in the “Discussions” section. Therefore, I think it would be good to add a second paragraph to section 4.3. In this paragraph, species of the mastixioid flora that have the same circumstances that are later mentioned in conclusions would be mentioned as an example:

their affinity for highly humid environments and “[Their] seeds are sensitive to drought, these factors may hinder their population regeneration. Given [Their] limited dispersal ability, habitat fragmentation, and anticipated aridification.” It is preferable to choose endangered species as mastixioid examples. After giving the examples, the paragraph can be finished by pointing out that there is a group of living mastixioid fossils whose survival is threatened by Climate Change, along with other drivers of Global Change.

Other minor improvements proposed are:

Line 53: please add the authorship, “Hand.-Mazz.” after the word “Diplopanax”

Line 70 to Line 72: It would be convenient to add the authors of the descriptions of the nine species mentioned. But it was previously suggested to redo the paragraph, so this would not make sense. In any case and for all parts of the paper, if a taxon is mentioned for the first time, the binomen is not enough, the authorship must be added.

Line 161: please change “combination [58].If there” to “combination [58]. If there”.

Line 281: please change “CO²” to “CO₂”.

Line 306: please add the authorship of the species Davidia involucrata.

Lines 320 to 321: Please add, in parentheses, the countries of those two locations.

Line 334: Please add authorship of the genera Metasequoia and Shaniodendron.

Line 336 to 337: Please reword the sentence as it does not make sense as it stands now. Indeed, between “higher latitudes in the north [88,89]” and “Alsophila spinulosa [90], and Metasequoia glyptostroboides [91]” there is a lack of a preposition or the link that relates both expressions.

Line 336 to 337(bis): Please add the authorship of the species Alsophila spinulosa [90], and Metasequoia glyptostroboides.

Line 375 to 376: Please omit the first sentence, as discussed above.

Line 386: Please change “Chin” to “China”.

Line 469: Please change “Hand．-Mazz.” for “Hand.-Mazz.”

Line 472: Please change “van der Bank” to “Van der Bank.”

Line 576: Please change “Methods in Ecology and Evolution” to “Methods Ecol. Evol.”

Line 608: please complete the reference as several data are missing.

Line 629: Please change “Prace Paleobot, 1975 Volume 4,” to “Prace Paleobot. 1975, 4.”

Line 675: Please change “Tree Genetics & Genomes” to “Tree Genet Genomes”.

Line 675: Please change “Philosophical Transactions of the Royal Society B-Biological Sciences” to “Philos. Trans. R. Soc. London. B Biol. Sci.”.

Author Response

Thank you for your question. We have made revisions to the manuscript based on your feedback. Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper entitled Distribution dynamics of Diplopanax stachyanthus Hand.-Mazz. (Mastixiaceae) and its implications to relict mastixioid flora conservation, deals with the modeling of a species of interest as Diplopanax stachyanthus Hand.-Mazz. which is a representative of the pluvisilvas of the infra-termotropical humid-hyper-humid pluvial and pluvial-pluvial-hyper-humid zones of Southeast Asia.

However, this manuscript does not provide any real new science, beyond validating a methodology that is more than commonly used in this type of study. There are countless publications that use the same methodological scheme of GBIF+WorldClim+MaxEnt, this scheme allows the publication of about 100000 papers on the distribution of species of vascular plants alone, without providing more science than a theoretical modeling of the distribution of a particular taxon without further ecological or biogeographical implications of the same.

In addition, the authors start from several premises that they assume to be true:

1) They assume that the GBIF citations are correct, and the taxa are correctly determined, or even that important citrons have not been lost due to GBIF filters.

Ref:

Beck, J., Böller, M., Erhardt, A., & Schwanghart, W. (2014). Spatial bias in the GBIF database and its effect on modeling species' geographic distributions. Ecological Informatics, 19, 10-15.

Zizka, A., Carvalho, F. A., Calvente, A., Baez-Lizarazo, M. R., Cabral, A., Coelho, J. F. R., ... & Antonelli, A. (2020). No one-size-fits-all solution to clean GBIF. PeerJ, 8, e9916).

2) They assume that the correlation between WorldClim bioclimatic data and actual data for their area, are correct. When in WordClim's own publication, the authors highlight examples of overestimation or underestimation of interpolations.

Ref:

Fick, S. E., & Hijmans, R. J. (2017). WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas. International journal of climatology, 37(12), 4302-4315.

Bedia, J., Herrera, S., & Gutiérrez, J. M. (2013). Dangers of using global bioclimatic datasets for ecological niche modeling. Limitations for future climate projections. Global and Planetary Change, 107, 1-12.

3) MaxEnt is a powerful tool, but it is a double-edged sword, it is very easy to use and implement, but it is very complex to understand how it really works, and to handle all the implicit variables that make the model, even being valid according to the AUC, to be really faithful to the distribution of a taxon.

Ref: 

Yackulic, C. B., Chandler, R., Zipkin, E. F., Royle, J. A., Nichols, J. D., Campbell Grant, E. H., & Veran, S. (2013). Presence-only modelling using MAXENT: when can we trust the inferences?. Methods in Ecology and Evolution, 4(3), 236-243.

Elith, J., Phillips, S. J., Hastie, T., Dudík, M., Chee, Y. E., & Yates, C. J. (2011). A statistical explanation of MaxEnt for ecologists. Diversity and distributions, 17(1), 43-57.

To increase the interest of the readers, in addition to increasing the novel scientific contribution, I suggest that the authors restructure their manuscript, provide new data on distribution, ecology, communities in which this taxon occurs, bioclimatology, biogeography, or focus the manuscript on the comparison of different distribution models. In short, that provide new science to work on in the future.

Therefore, I do not recommend its publication in Forest.

Some words are not well translated, or the ideas are not well understood.

Author Response

Thank you for your question. We have made revisions to the manuscript based on your feedback. Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

lines 61-65: Additionally to climate impact, does human activity influence distribution of Diplopanax stachyanthus negatively or positively? Particularly, could migration of this species to mountain regions (as said in line 333) be due to agricultaral development of plain regions in historical time?

lines 60-61: photo of Diplopanax stachyanthus living plant would decorate the article. :)

lines 123-129: The said averaging of climate models should be explained in more details. Firstly, whether the calculated climate values (temperature values etc.) were averaged or model parameters? Secondly, what was the benefit obtained after averaging? Thirdly, did you come up with the idea of averaging yourself?

line 134, Table S3, Table 1: In an ecological study, the value 0,8 as the lower threshold of confirmed correlation is too strict criterion (and your previous publication is not a compelling argument to apply this threshold; at least, an independent reference should be demonstrated). To justify correlation analysis and variable selection, necessary is to place all paired significance values in Table S3 (into its upper right area); after this, preferable is to select variables those correlate non-significantly. For example, Ø. Hammer's PAST soft package (free for download) gives a quick way to calculate significance values for all pairs simultaneously and arrange them in a table.                                          As for your selected 7 variables, it is obvious that bio2 and bio3 strongly correlate (simply, because of bio3 is calculated through bio2). And Table S3 demonstrates that two temperature parameters (bio8 and bio10) also correlate strongly (perhaps, because bio8 is used in calculation of mean annual temperature bio10 and strongly influences the last), and two precipitation parameters (bio13 and bio18) do the same (because the warmest quarter includes the wettest month, as it is said in line 293). At first sight, the appropriate combinations of weakly correlating variables could be as following: 1) bio19 & bio8 & bio13, 2) bio18 & bio2 & bio10. At this, the fewer variables the model includes the bigger contirbution each of them gets. So, imperatively recommended is to analyze, at least, the model including only the revealed variables of most contribution (bio2, bio18, bio19; according to Table S3, they do not correlate very strong), whereas the two combinations proposed or any other ones also may be analyzed if interested. 

lines 141-143: "Model calibration training is conducted for all models with 70% of the distribution data, as the test set for cross-validation during calibration, allocate the remaining 30%, with 10 replicates and cross-validation as the run type." From this phrase, it is not clear enough (maybe, the only because of not quite correct grammar) whether 10-fold cross-validation was done with the training data (70%) or with the test data (the rest 30%; cf. with lines 155-157 where the sense is completely clear). The first way (within the training 70% of data) is correct because cross-validation is a step of calibration, and after this, further independent validation of the model is necessary to do with the rest 30% of distribution data, evaluating the real accuracy.

lines 161-162, Figure S1: Discussed in a recent publication (see reference further) is that AIC is not a suitable criterion to discriminate between good and non-good species distribution models (Velasco, Julián A., and Constantino González-Salazar. "Akaike information criterion should not be a “test” of geographical prediction accuracy in ecological niche modelling." Ecological Informatics 51 (2019): 25-32)

lines 171-172, 193-196: "Model performance was validated using cross-validation methods." Despite the word "validation" included in its name, cross-validation is not a method of model accuracy evaluation, but the only additional way to optimize a model. So, cross-validation applied when building the final model looks strange.

lines 180-181: "... converted Maxent's predicted map into a binary grid layer (unsuitable: 0; suitable: 1)". Which was suitability threshold applied to the value 1?

Figures 2 & 3: Visually, very similar are the maps built for LGM cold time (Fig. 2a) and future warm decades in some prognoses (Fig. 3e, i, k, l). So, it is worth to discuss in more details what is the sufficient climate similarity between LGM time and considered future scenarios. 

lines 290-294: To make a reader's understanding clearer, it is worth to place the climate description into Section 2.1.

Notes on technical detailes:

line 30: Mean Diurnal Range of what? Here, the explanation should be given with the first mention of this variable in the text.

line 84: Written first time in the text, the abbreviations LGM, MH should be explained (not only in the abstract). On the other hand, since the two are explained in the text below (lines 118-119), the terms may be written here in their full form.

Figure 1: On this map presenting the initial data, the colored areas that are part of results (as were constructed by the applied model) should be removed.

line 197: An especial map of modelled distribution should be placed in Section 3.1.

Table S2: Titled "Percent contributions of the bioclimatic in the species distribution models for Diplopanax stachyanthus", the tables containes only variable names and measure units, and doesn't contain values of their contributions to any models.

Figure S2: In the figure cap, explanation should be given that, according to lines 190-192, presented is the final built model.

Figure 6: The visual difference between solid lines of the same color is not obvious; in each pair one line needs to make thicker. 

lines 411-412: Figure S3 does not show Receiver operating characteristic curve.

lines 414-415: Table S4 does not contain the results of correlation analysis.

lines 141-143: "Model calibration training is conducted for all models with 70% of the distribution data, as the test set for cross-validation during calibration, allocate the remaining 30%, with 10 replicates and cross-validation as the run type." From this phrase, it is not clear enough (maybe, the only because of not quite correct grammar) whether 10-fold cross-validation was done with the training data (70%) or with the test data (the rest 30%; cf. with lines 155-157 where the sense is completely clear). 

Author Response

Thank you for your question. We have made revisions to the manuscript based on your feedback. Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Dear authors,

I would like to congratulate you with a well-written study. Clarity of the text was a considerable improvement in comparison to most other manuscripts I reviewed in last years by authors from your country. I also find the reported results interesting and potentially useful for the broad community of conservationists. I am not qualified to comment on methodology in detail, but my general impression is that the methods used in the study are robust and mostly well justified. Therefore, in my view your study should be accepted for publication, provided you will properly respond to my comments below.

The Introduction reads well, but justification for using the study system is rather weak. The justification should not be based on a shaky basis that the species is supposedly “famous”. There are hundreds if not thousands of Tertiary relict species in southern and eastern China, and many of them are more endangered than the species you have chosen. So why did you decide to study Diplopanax?

In Discussion, you refer to a study of a closely related species. The study (Tang et al. 2017) was based on mostly the same concepts and methodology, so its results are highly relevant and should be mentioned already in Introduction. Another highly relevant study (of the same species as you have selected) was focused on phylogeopgraphy in addition to the niche distribution modelling (Feng et al. 2019). It should also be mentioned in Introduction. Referring to these studies in Introduction will require a better explanation of what we already know about the selected study system, and what kind of potentially useful information is missing. Your current introduction of systematic issues with placement of the genus might be useful but it was not especially relevant to the aims of your study. At the same time, the lack of introduction of highly relevant results on the same or closely related species is not acceptable.

The same applies to the lack of report on basic data used in you study. You mention gathering of 379 distribution records for the species. The proper practice would be to report all these records according to the standards of GBIF. Each record should report GBIF ID number, a link to the original database of the record (preferably on-line, if possible), the deposition numbers for herbarium specimens, names of collectors, date of collection, name of locality and province, coordinates, altitude, and (very important!) approximate error of geolocation. The errors can be very different for some old records from 19^th century, which often mention only country or province of collection, and the contemporary GPS-geolocated records with precision of few meters.  Without knowing the levels of precision of geolocations used in your analyses it is impossible to evaluate the quality of your niche distribution modelling.

I would recommend filtering out closely located distribution points after a preliminary selection based on quality of geolocations. GPS-geolocated records should be all preserved together with other records of high quality (errors less than 100-1000 m). These records may be supplemented by less accurately geolocated records (up to 10 km), provided they are not present in the same 0.2° × 0.2° grid with localities of precise geolocation.  Filtering out of closely located distribution points should be based on this adjusted filtering scheme. Selected for both analyses (before and after the filtering out) records should be highlighted (as italics and bold font, for example), but they should be reported together with all other basic records in the same table.

I understand that my recommendation bears the risk of considerable additional working load for you. Nevertheless, failing to report the levels of precision for your records and a lack of concern about the level of precision of your basic data cannot be accepted.

Author Response

Thank you for your question. We have made revisions to the manuscript based on your feedback. Please see the attachment.

Author Response File: Author Response.pdf