Subtle East–West Phylogeographic Break of Asteropyrum (Ranunculaceae) in Subtropical China and Adjacent Areas

Round 1

Reviewer 1 Report

I found the manuscript well-written, except for a few issues that you can easily attend to. The manuscript is interesting and the results based on chloroplast and nuclear markers are pretty good. I attach the manuscript with a few comments. 

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

 

Thank you for the comments about the manuscript “Subtle east-west phylogeographic break of Asteropyrum (Ranunculaceae) in subtropical China and adjacent areas”. According to your suggestions, we have revised the manuscript carefully and the modified sentences were marked in red color. The following lists the revisions that have been made in the revised manuscript.

Response: There are multiple questions for some sentences in Page 3, and we would change them in revised manuscript.

Nuclear marquers usually have double peaks (polymorphis) do you get double picks? If you do, how did you treat them?

Response: We chose ten single-copy nuclear loci, so the haplotype phase of them could be determined by applying the phase algorithm in DnaSP 5.10.

Which groups? You need to explain which are you groups, and what was the decition to name it.

Response: We named two groups based on the structure result, of which Group_1 included MT, PT, GS, FG, ZX, YL, EMS, ELS, DJY, PL, SNJ, HZG, LCG, WLS, while Group_2 consisted of MJ, SJ, LB, MB, WL, YY, LS,YS, XN,ZY, HP.

Which is the frecuency of each haplotype? Number or individuals?

Response: There are 46, 7, 22, 84, 21, 54, 24, 19, 21, 10 individuals from H1 to H10, respectively.

Did you find recombinant nucleotides? Did you eliminate them from the analysis?

Response: Yes, we indeed found recombinant nucleotides in several nuclear loci, and eliminate them with linkage disequilibrium sites from the following analysis.

What does mean “were combined”?

Response: The word “combined” maybe not very suitable, so we changed it to “jointed”.

I think you want to mean groups instead eigh nuclear loci?

Response: Yes, thank you for your reminder, and we have changed it as two groups and Asteropyrum.

These two clade are totally arbitrary.

Response: We have described them in the caption of Figure 1 and revised this part in revised manuscript

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript “Molecular phylogeography of Asteropyrum (Ranunculaceae) in 2 subtropical China and adjacent areas: Mid-Pleistocene differentiation and secondary admixture during the Large Interglacial (500-460 ka)” by Cheng et al. provides a nice overview of the biogeography of a perennial herb with a controversial taxonomic history. Despite not using a lot of data, the authors manage to reach solid and sound conclusions. The presentation of the analyses done is accurate, and nicely framed by a good introduction and well written discussion. I believe that some analyses presented in the paper could have done in a simpler way, but this does not affect the quality of the paper.

For instance, the scenarios explored in the ABC are quite redundant. First, it is not obvious why running different sets of simulations for data excluding or not the “admixed group”: first, knowing that admixture occurred between the two populations anyway make irrelevant the time estimates in the model with two populations; second, since the authors seem to have power to estimate parameters even in the more complex model, why simplifying it so much? Third, I do not think it makes much sense to include four different scenarios in the model with two populations: population growth and changes in Ne would have been better modeled as a continuous parameter, without any contrast between discrete models. Note in fact that here there is no real contrast between competing models. Fourth, the “Structure” analysis is very explicit: the “mixed” group totally looks like admixed. Now, I agree it is always better to have more than one independent validations of a result, but definitely some of the models tested were irrelevant and did not need to be tested. Something which is not entirely clear to me from the paper, and I urge the authors to explain better, is how the mutation rate was estimated. From the text it seems it was estimated, but its prior and estimate is not presented anywhere. For instance, where is mu in Figure S1 or Table 4. In section 3.3 there is some explanation, but I think it is not clear, and since mutation rates can have linear impacts on Ne and time estimates, this should be clarified. One more small thing: I think it is inaccurate writing just “Probability” in Figure 4 for the posterior probability densities.

Concluding, I think that after some minor adjustments the manuscript is ready for publication.

Author Response

Dear Reviewer,

Thank you for the comments about the manuscript “Subtle east-west phylogeographic break of Asteropyrum (Ranunculaceae) in subtropical China and adjacent areas”. According to your comments, we have revised the manuscript carefully and the modified sentences were marked in red color. The following lists the revisions that have been made in the revised manuscript.

For instance, the scenarios explored in the ABC are quite redundant. First, it is not obvious why running different sets of simulations for data excluding or not the “admixed group”: first, knowing that admixture occurred between the two populations anyway make irrelevant the time estimates in the model with two populations; second, since the authors seem to have power to estimate parameters even in the more complex model, why simplifying it so much? Third, I do not think it makes much sense to include four different scenarios in the model with two populations: population growth and changes in Ne would have been better modeled as a continuous parameter, without any contrast between discrete models. Note in fact that here there is no real contrast between competing models. Fourth, the “Structure” analysis is very explicit: the “mixed” group totally looks like admixed. Now, I agree it is always better to have more than one independent validations of a result, but definitely some of the models tested were irrelevant and did not need to be tested.

Response: Thanks very much for your nice comments. I have deleted the scenarios with two populations in the manuscript according to your suggestion.

 

Something which is not entirely clear to me from the paper, and I urge the authors to explain better, is how the mutation rate was estimated.

Response: Firstly, two mutation rates were chose as candidate rate based on rates of herbaceous angiosperms estimated by Richardson et al. (2001). After two ABC simulations (two populations and three populations), we found the estimate mutation rate were both close to 3.46×10^-9. So we chose this rate to estimate the divergence time finally. The Bayesian skyline plot of 8.69×10^-9 in Figure 3 maybe cause misunderstanding, so we would delete it.

From the text it seems it was estimated, but its prior and estimate is not presented anywhere. For instance, where is mu in Figure S1 or Table 4. In section 3.3 there is some explanation, but I think it is not clear, and since mutation rates can have linear impacts on Ne and time estimates, this should be clarified.

Response: Thank you for your comment. We have added the priors for all estimated parameters in supplementary Table S2.

One more small thing: I think it is inaccurate writing just “Probability” in Figure 4 for the posterior probability densities.

Response: Thank you for your advice. We would change it as posterior probability density distributions.