Deep sampling of Hawaiian Caenorhabditis elegans reveals high genetic diversity and admixture with global populations

Essential revisions:

Drawing from reviewer 1 we all agreed that the paper should acknowledge the current limitations of sampling more, i.e. it may be that the Pacific region at large shows a high diversity and there is nothing special about Hawaii.

Yes, we completely agree about this concern. In the Results (“Hawaiian C. elegans are divergent from most strains sampled across the globe” and “C. elegans population structure”) and the Discussion (“The origins of C. elegans”) we have added content to address the sampling biases in Hawaii and Europe and more systematically indicate that we cannot know whether this diversity is specific to the Hawaiian Islands.

The reviewers' full reviews are given below. The points they raise are reasonable, and I think can be easily be addressed mostly just with some rewriting.

Reviewer #1:

[…]

A few suggestions to improve the manuscript:

1) It may be that the Pacific region at large shows a high diversity and there is nothing special about Hawaii.

The present sampling is on the Hawaiian islands. Prior sampling is mostly in North America and Europe, and little in the Pacific region, which introduces a bias. This should be made clearer, for example in the second paragraph on subsection “Hawaiian C. elegans are divergent from the global population” and in the Discussion. The Discussion is centered on the center of origin of C. elegans and the supposed ancestral niche. Although the Pacific region is mentioned, it would be good to more systematically indicate that one cannot distinguish at the moment whether this diversity is particular to the Hawaiian islands. That C. inopinata was found in South Japan could be mentioned.

Thank you for pointing out this limitation. In the Results and the Discussion we have added content to address the sampling biases in Hawaii and Europe and more systematically indicate that we cannot know whether this diversity is specific to the Hawaiian Islands.

Second paragraph of subsection “Hawaiian C. elegans are divergent from the global population”: note that CeNDR shows more strains with the corresponding "Hawaiian-like" isotypes (BRC20231 from Taiwan (?), JU1171 from Chile, thus from the Pacific region). In line with this, it could be noted in the penultimate paragraph of the Results section that the admixed isotype was QG556 (and QG558), from California, also on the Pacific rim.

These are excellent points. We have reworked the manuscript to highlight the Pacific region isolations for the MY23 isotype. In paragraph three of subsection “C. elegans population structure” we added this detail about QG556 and a new figure (Figure 6—figure supplement 1) that more clearly shows admixture and geographical proximity between the Hawaii Invaded and non-Hawaiian C populations.

2) Could it be that the relatively high diversity is due to a lack of selective sweeps reducing the diversity (due to the high linkage disequilibrium) in the Hawaii set because of geographic isolation? Diversity may have been higher elsewhere before and lost perhaps due to human activity and movement.

We added this possible explanation for the high genetic diversity found among the Hawaiian isotypes (subsection “The origins of C. elegans”).

3) Gene flow between "Hawaii/Pacific" and the rest of the world.

I do not understand why there is emphasis on the gene flow with the C group and little is made (i) of the gene flow to the G group that is shown in Figure 5—figure supplement 2 and Figure 7—figure supplement 1; and (ii) of the admixture with the “HI Low” of some non-Hawaiian strains (including the G group). It seems that both “HI Low” and “HI Invaded” are partially admixed.

Thank you for this valuable feedback. We have added text to the Results (subsection “C. elegans population structure”) to reflect the evidence of gene flow from a population related to the Hawaiian Invaded and Hawaiian Low populations to the non-Hawaiian G population. We have also added to the Results (final paragraph) regarding the haplotype sharing between the Hawaiian Low and non-Hawaiian populations. Furthermore, we specifically mention that the Hawaii Low population contains some admixture with non-Hawaiian isotypes and shows some evidence of globally swept haplotypes (subsection “The ancestral niche of C. elegans might be similar to the Hawaiian niche”).

4) The category of “fruits, nuts or vegetation”, opposed to “flower” is odd. First, nuts are fruits. Second, flower is vegetation. Please at least change the name and perhaps reconsider the categories.

We agree that the substrate category names in the original manuscript were potentially confusing. We have changed the names and reorganized the samples into new categories. After changing the names and categories, we reran all statistical calculations involving substrate categories.

1) We removed the eight “vegetation” samples from the original “fruits, nuts or vegetation” category and recategorized them into a new “vegetation” category.

2) We recategorized all the samples originally called “other” into the new vegetation category because these samples all consisted of various forms of vegetation that were each infrequently sampled.

3) We changed the name of the original “Fruit/nut/veg” category to “Fruit” because nuts are fruits, as pointed out.

We now include our logic for choosing the six major substrate classes in the Materials and methods section: “On these hikes, we opportunistically sampled substrates known to harbor C. elegans,including fruits, flowers, stems, leaf litter, fungus, compost, wood, live arthropods, and molluscs (Ferrari et al., 2017; Frézal and Félix, 2015; Schulenburg and Félix, 2017). We chose to group these substrates into six major categories: leaf litter, fruit, flower, fungus, invertebrate, and vegetation. The vegetation category contains all substrates that do not fall into any of the other five categories. We chose to distinguish the leaf litter, fruit, and flower substrates from the vegetation category because we suspect these substrates provide distinct microhabitats and each substrate was sampled extensively.”

Importantly, while we were updating the categories, we discovered 19 samples that had been inadvertently filtered from our data set prior to plotting Figure 2 and running the statistics involving substrate categories in the original manuscript. These samples have been added back into the data set for plotting and analysis. No inferences from our statistical tests changed because of the reorganization or the reintroduction of the 19 missing samples.

5) I could not find the correspondence between the samples, the strains, the species, the isotypes. The supplementary table with the sampling detail lacks the strain and species names in each sample.

The sample, strain, species, and isotype names are all included in Source data 1. The column names for each are; c_label, strain, species_id, and isotype. We have updated the text to reference Source data 1 (full collection data) specifically, rather than just Supplementary file 1 (table of collection categories by island), which might have been confusing.

6) The Discussion is currently centered on the center of origin of C. elegans. However, some other points could be discussed, for example the use of these data and strains as resource, e.g. for GWAS, the sampling design, the other species, etc.

We agree that the utility of these data in GWA should be addressed in the discussion. We added text within the first paragraph of the Discussion to highlight both the utility and potential disadvantages of these data with respect to GWA (First paragraph).

Reviewer #2:

[…] My criticisms are mostly to do with wording, specifically around the meaning and consistency of use of "population."

Subsection “Hawaiian C. elegans are divergent from the global population” and elsewhere: use of the word "population" to refer to the Hawaiian vs. non-Hawaiian samples seems a bit inconsistent/misleading, as the additional analyses show that they each comprise multiple genetic clusters (also referred to as populations) and the non-Hawaiian sample in particular is geographically global, and so may confuse readers. I would be more comfortable with "population sample" or "sample of strains" or just "sample." This issue appears again in paragraph two of the Discussion after much effort went into analyzing distinct genetic groups within the Hawaii sample that are not entirely defined by geography. In subsection “The ancestral niche of C. elegans might be similar to the Hawaiian niche”, it seems the use of "populations" might better be replaced by "localities." Please comb through the manuscript for use of "population" to ensure consistency in what is meant, and what the analysis shows.

Thank you for this helpful criticism. We revised use of the word “population” throughout the manuscript to avoid confusion and ensure consistency with the results of our analyses. We no longer use the word “population” to refer to the Hawaiian and non-Hawaiian isotype samples; we now refer to them as “samples”. The term “Hawaiian population” has been removed. We also removed the use of the word “populations”. We changed this to “Hawaiian sampling locations”.

Subsection “C. elegans population structure on Hawaii”: I think this conclusion may be too strong. The admixture analysis does provide evidence for 11+ genetically-distinguishable groups, though whether these necessarily correspond to (geographic) ancestral populations is tricky to infer. This challenge is especially true given the highly selfing reproduction of C. elegans that violates assumptions for what populations mean in admixture, so some caution is needed in interpretation. I'd be more comfortable with wording along the lines of, "This analysis identified at least 11 genetic groups for this sample of C. elegans, each of which we refer to as a population for the purposes of further analysis."

We thank the reviewer for this helpful comment. We have updated the admixture Results section (subsection “C. elegans population structure”) to include the suggested text. Throughout the manuscript, we are consistent with our usage of the term “population” to refer to genetically distinct groups identified by ADMIXTURE.