Strong Population Genetic Structure for the Endangered Micro-Trapdoor Spider Moggridgea rainbowi (Mygalomorphae, Migidae) in Unburnt Habitat after Catastrophic Bushfires
Marshal Hedin
Round 1
Reviewer 1 Report
The manuscript “Strong population genetic structure for the endangered microtrapdoor spider Moggridgea rainbowi (Mygalomorphae, Migidae) is well-written and contains relevant information on an endangered trapdoor spider from Australia. Authors find very obvious population structuring in Moggridgea whose populations probably represent evolutionary significant units, relevant for biodiversity research and subsequent conservation planning.
While molecular analyses have some weaknesses, namely a relatively low number of samples, a combination of older (2013) and fresh (post-2019) material for population genetics, and relatively low information value from COI and ITS genes however, everything is clearly described and justified. Also, having an outgroup for the phylogenetic part would be informative.
The manuscript (especially the methods) is very detailed, so I do not see any significant problem. I would recommend slightly improving:
Figure 1 – I couldn’t immediately find Kangaroo Island, both greens are too similar. I suggest you encircle KI.
Line 107: Please check if the ethanol was really 100%.
Line 259: On which base did you decide to root the trees to the western clade? In this case, for example, having an outgroup would be useful.
Figure 7 – should have more explanation, to make it more or less self-explanatory. What are the values in the legend, do colors mean high, medium, or low fixation or a specific number etc.
Kind regards
Author Response
Please see attachment
Author Response File: 
Author Response.docx
Reviewer 2 Report
Excellent and important paper, thank you. Some of my suggestions below are trivial, others would work to strengthen an already strong paper
Line 54 – “first such case recorded”. Oceanic dispersal has been previously hypothesized for Pacific Island barychelids, so perhaps change word choice for “recorded”?
Line 107 – Perhaps clarify how many animals were sacrificed, versus those released after removing a leg? Is it known if the spiders survive after leg removal?
Line 226 – Not sure about neighbor-joining here? Also, why not SNP direct inference, e.g., using SNAPPER?
I’m not sure that FIG 3 is needed a part of the paper? Perhaps move to Supplemental?
I wonder why the authors did not calculate allelic richness, heterozygosity, and Fst from their SNP data? The former measures would also be quite interesting / informative from a conservation perspective.
FIG 4 – Include a map inset?
FIG 5 – I would recommend using a color other than grey for the 3rd cluster. Almost appears as missing? Also, why not array populations from west to east, geographically adjacent pops adjacent on the STRUCTURE plot?
Line 407 – I would not infer male-based gene flow based on ITS, since the nuclear SNPs didn’t suggest this pattern
Line 418 – perhaps show this historic break on the map?
Author Response
Please see attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
This manuscript aims to report a strong population genetic structure of an endangered micro-trapdoor spider Moggridgea rainbowi, on Kangaroo Island, Australia. The data presented in the manuscript demonstrated the critical information that would certainly contribute to conservation decisions for this endemic spider with only a few populations. The population genomic data generated from the restriction-site associated DNA sequencing (e.g., the SNP data for thousands of genetic markers)did illustrate the substantial population subdivisions with this species. The analyses for fixed alleles in each population highlighted the genetic uniqueness of each genetic pool in different geographic areas. In these regards, this manuscript has its scientific significance to be published in diversity. However, I have several concerns before this manuscript can be formally published.
The authors should put this study in a hypothesis-testing framework. Instead, for the population genomic data and Sanger data, the authors only conducted the descriptive presentations and analyses for the data, which the important explanations can only be inferred in the discussion. The authors conducted the population genomic analyses (STRUCTURE, DAPC, PCA, and pairwise comparison of allelic fixation between populations. These analyses should be considered as the preliminary analyses of the genomic data. The tree reconstructions using mitochondrial COI, nuclear ITS, and SNPs conflict with each other. I would suggest conducting more advanced analyses on gene flow, GWAS, and divergence time estimation to give a more comprehensive understanding of the gene pool conductivities among populations, the local adaptations of each population, and the possible geographic history responsible for the population differentiation. In the discussion, the authors already provided a good review of the geographic histories that could be responsible for the population subdivisions, which could be a set of reasonable candidate hypotheses. These more advanced analyses should be considered in the next revision. The discussion should be revised according to the new analyses and results.
The authors presented eight figures in this manuscript, which generated more confusion than an explanation to the readers. Figure 1 is a map (without labeling the localities of the collection sites on the map, but in the figure legend), and Figure 2 is the photos of the spiders. These two figures can be merged into one to save some space. Figure 3 is not a result of this study. It is quality control of the bioinformatic treatments that belong to the supplementary materials. Figure 4 presents the result of PCoA. Within the same space, it would be good to include the result of DPAC, a method that reports results of k-mean clustering, thus, is more informative for comparison. Figures 5 and 6 report the STRUCTURE results of different levels. Figure 7 reports the allelic fixation comparisons between populations. These three figures should be on one plate so it would be easier for readers to compare and identify which population has its own ancestry and how many fixed alleles occurred in this population. The results of delta K should be moved to supplementary materials. This new figure, should include results of the gene flow estimations between populations. Figure 8 shows the phylogenetic/phylogenomic trees constructed from COI, ITS, and SNPs. Which one is considered the preferred tree? If the preferred tree is decided, the divergence time should be estimated to accept or reject the hypotheses of which geographic historical event was responsible for the population subdivisions of this spider. In general, the font sizes of the labels, axis, and titles are nearly invisible to the readers in all figures, which needs to be fixed.
The specific comments are listed below:
Line 17. Use "population genetic structure" instead of "genetic population structure."
Line 22: Provide numeric results here instead of descriptive statements.
Line 23. Provide the estimated divergence time.
Line 56. A species that is closely related to M. quercina group, an African species.
Line 94. This sentence is long and hard to read.
Line 102. Genetic diversity, therefore, often the population level, is the focus of conservation plans, which has been highlighted as one of the three most essential biodiversity levels for conservation. When discussing the conservation plan, there is no need to suggest the importance of knowing the population's genetic structure. Instead, the authors should highlight what hypotheses they want to test in the study.
Line 108. Use September 2nd, 2019 .... to avoid confusion.
Line 113. Separate "we needed ..." into another sentence.
Line 114. State the duration of life span here to replace "long lives."
Line 118. "to reconstruct phylogeny". Not to "produce".
Line 120-123. The orders of the localities in Figure 1 legend, table 1, and text are all different, which makes it hard to read. Using only one order is suggested.
line 126. “....green.”, “.” is missing.
Line 127. Are there 'East' groups? If yes, which ones are considered east?
Line 127-129. These localities should be mapped on the map (B)
Line 132. The figure legends are inconsistent in styles.
Line 145. Usually, it is written "Restriction Site Associated DNA sequencing."
Line 153. State the name of the Illumina Kit.
Line 180, 182, 184. Use formal symbols.
Lin 190-192. Please explain how to get a no-missing data matrix. It is unclear how removing individuals with missing data >70% and singletons would generate a 100% complete matrix.
Line 192-193. Please explain why n<15 was considered as a low sample size.
Line 262. Italicize the species name.
Line 282. This is a different style than in methods.
Line 307. I cannot find a red dot anywhere. The color for a population should be the same across different figures.
Line 311. Table S1 and Figure S3 only have the voucher numbers; I believe the readers would have difficulty telling the localities.
Line 322 to 326. This figure is nearly impossible to read. The font sizes on the axis are way too small to tell. The dots on the curve in (a) is almost invisible.
Line 327. Eastern KI samples were not defined. In addition, because the font sizes in figure 6 are too small, it is difficult to capture the referred populations.
Line 331. Pen/ Penneshaw were used alternatively in figure legends and text, which generated confusion for readers.
Line 333. There is no "grey" population in Figure 6b. please clarify.
Line 351. P and p were used alternatively in text and figure legends.
Line 353. Suggested editing. "..., indicating gene flow occurred between these two localities." Gene flow does not connect populations.
Line 357. The authors included the results of IQ tree from the SNP data, which should be reflected in the subtitle.
Line 358-361. If the authors did not include the confounded sequences in the final analyses, this part should be removed.
Line 374-383. I am not sure of the purpose of this result (Figure 8). The COI result would be expected to be different from ITS result because they are different genes with different evolutionary rates and histories. The typical way to deal with the multiple loci Sanger data set is to concatenate the gene sequences. However, the author seemly intends to compare the results of these two genes to the results of SNP data. Obviously, the SNP data outperformed the Sanger data judging via the nodal supports. In this regard, the Sanger data just did not provide good-enough informative phylogenetic signals when compared to SNP data. Additionally, it would be good if the authors conducted a divergence time estimation analysis.
Line 397-398. This is not a valid statement. The ITs result has very low support for the authors to claim any conclusion.
Line 401-403. I believe there is low gene flow among those populations as stated by the authors. However, having fixed alleles in the population (in this case, it is pairwise differences of unique loci, not alleles), is not a direct estimation of gene flow. There are cases that gene flow still occurs after speciation. The authors should conduct a gene flow estimation instead of inferring low gene flow by the observed number of fixed loci.
Line 408. Since there is already a reference for time calibration, the authors should conduct a divergence time estimation for the ESUs in this study.
Line 416-418. These all provide good time calibration reference points to conduct hypothesis testing for identifying which geographic event is most likely responsible for the population differentiation of these spiders.
Line 473-480. It is better to draw a conclusion from the results of this study. Making a general statement about the importance of population genetic study for conservation plans that is well-known is unnecessary. Pointing out how the results of this study, e.g., the application of population genomic data to reveal the cryptic diversity and thus to make advisory suggestions for the conservation of these endemic spiders, is the significant contribution of this study.
See above
Author Response
Please see attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 3 Report
This is the second time I review this manuscript. I found the revised version is improved. I only have few more suggestions in figures (see below). In the reply to reviewer document, the authors explained the reason they did not conduct a divergence time estimation analysis using their current SNP data. I still feel using the time calibrated tree reconstructed from SNP data to conduct the further analyses, such as DIYABC, for testing the demographic or other population genetic dynamic history and their relevant hypotheses should be considered in the future. However, as the authors stated, they have low number of samples from an endangered species, they would remain their aim of this study is to illustrate the population genetic structure and the genetic clusters for conservation units. I agreed this is an important result, even without further hypothesis testing. According to this reason, I would agree this manuscript could be published in diversity as the current results of analyses.
Several suggestions on figures.
Figure 1. Figure 1b should have a scale bar to indicate the size of the spider.
Figure 3. Increase of font size of the axes of figure 3 a and c. Also, I would check the data quality of the of the individual with multiple ancestries in the “west” population just to confirm the multi-ancestry pattern was not caused by low quality of the data).
Author Response
Please find our response to the second round of comments from Reviewer three:
Point 1: This is the second time I review this manuscript. I found the revised version is improved. I only have few more suggestions in figures (see below). In the reply to reviewer document, the authors explained the reason they did not conduct a divergence time estimation analysis using their current SNP data. I still feel using the time calibrated tree reconstructed from SNP data to conduct the further analyses, such as DIYABC, for testing the demographic or other population genetic dynamic history and their relevant hypotheses should be considered in the future. However, as the authors stated, they have low number of samples from an endangered species, they would remain their aim of this study is to illustrate the population genetic structure and the genetic clusters for conservation units. I agreed this is an important result, even without further hypothesis testing. According to this reason, I would agree this manuscript could be published in diversity as the current results of analyses.
Response 1: We thank Reviewer Three for their constructive and insightful comments, which we agree have strengthened the paper
Point 2: Several suggestions on figures.
Figure 1. Figure 1b should have a scale bar to indicate the size of the spider.
Figure 3. Increase of font size of the axes of figure 3 a and c. Also, I would check the data quality of the of the individual with multiple ancestries in the “west” population just to confirm the multi-ancestry pattern was not caused by low quality of the data).
Response 2: We have made the suggested changes, adding a scale bar to Figure 1 and increasing the font sizes of Figure 3.
