Hello Diogo,

Greetings from France.

Sorry for the delay, here are my answers to the student questions:

Questions directed to Jonathan Romiguier, about the paper ÒComparative population genomics in animals uncovers the determinants of genetic diversityÓ.

>1. (Group 5N).  Do you think the difference in timescales of the historical contingency events can affect their impact as studied in the article? We ask this because some species could have experienced events on a scale of a few decades (e.g., conservation and invasive species status) and events involve longer timescale (e.g. bottlenecks or species range). These scales are so different that it may be reasonable to expect them to have different impacts on observed diversity (e.g. ancient and long-lasting events having greater impact). Moreover, do you think the proxies for distant events are as reliable as those for more recent history?

The difference in timescales of the historical contingency events should probably impact differently the observed diversity. The analyses of this article did not detect any effect of historical contingency on genetic diversity, but the problem might be that proxies for historical contingency are just difficult to identify. However, if we can test this better one day, my bet would be that only long and pervasive timescale events can impact observed diversity. In my opinion, if every little ecological disturbances could affect durably genetic diversity, it would have been close to impossible to predict genetic diversity from life-history strategies. It doesn't mean that short-term ecological disturbances cannot have dramatic impact (such as leading to the exctinction of a species), but just that we should not expect to see every ecological disturbances translated into genetic diversity declines.


>2. (Group 2N). Your study was made using data from groups over many animal taxa, and showed that variability depends on the life strategy. Do you think the same pattern will occur within a group of more closely related species?  Take insects as an example, in which we have the r/k diversity necessary for the study within a more closely related group (recalling insects have a broad range of strategies). The idea is that with phylogenetically related species, we would remove more confounding factors not related to historical contingency or life-history.

I do think that the same pattern will occur within a group of more closely related species. We did not have the possibility to give more details in the article (space constraints), but you can see interesting contrasts among closely-related species in this dataset. By example, in the article, if you take only the insects, you have a clear effect of the eusocial lifestyle (K-strategy: ants, bees, termites) vs solitary lifestyles (r-strategy: butterflies, mosquitoes). You can read more about the effect of the eusocial life strategy on molecular evolution in the following article: https://www.ncbi.nlm.nih.gov/pubmed/26227898 . There are other striking non-insect examples: both in urchins and Nemerta, we have a clear effect of brooding species (K-strategy) vs species with external fecundation (r-strategy).

>3. (Group 2N) Considering the known differences between asexual and sexual animals (in particular the accumulation of deleterious variants and lack of recombination in asexuals) what do you think a similar study but involving asexual strains could tell us? Has such an approach been attempted?

There are many specific patterns that we expect in asexual animals (mostly related to the loss of recombination): reduced effectiveness of selection, changes in the amount of neutral polymorphisms segregating in populations and an arrest of GC-biased gene conversion (which is a process that increases the GC-content of a genome). What I would expect in a similar study involving asexual strains is that asexual animals (if asexuality is sufficiently old in their lineage) would be outliers regarding their life-history traits. Such studies have been rarely attempted, but you can read here a very recent and relevant paper about asexuality and molecular evolution in stick insects : https://academic.oup.com/mbe/advance-article/doi/10.1093/molbev/msy058/4962172

>4. (Group 9D) Are there evolutionary implications for K strategists of having higher Êpi_n/pi_s ratios? Considering a scenario where these strategists go through a bottleneck, could this difference imply more risk of extinction for these species (for example, because they would already have an increased amount of deleterious mutations in the first place)? Or do you think the non-synonymous nucleotide diversity wonÕt affect this risk because the mutations acquired are only weakly deleterious?

It's very difficult to answer to this question and I doubt that an universal answer (i.e. which would be true for every cases and species) exists. It's more than likely that in some cases, the accumulation of slightly deleterious mutations is a burden that might be a problem in case of bottlenecks, but the real question would be more about how frequent are such cases. According to an interesting theoretical study (http://www.jstor.org/stable/pdf/2462976.pdf ), effective population size lower than 100 are likely to have an increased risk of extinction.

>5. (Group 8N) Is it possible that the higher variability for some species may make them more likely to occupy new niches and to differentiate themselves, being more susceptible of evolutionary processes that would form new species? Could such an explanation elucidate the fact that invertebrates, particularly arthropods, are one of the most diverse species groups? In other words, do the processes which drive variability within species also potentially provide the basis for the differentiation related to speciation? A possible test for this hypothesis would be to investigate if speciation rates (or phylogenetically defined diversification) is related to properties on the r/k spectrum.

It's a common hypothesis in the scientific litterature, but I personally think that higher variability and speciation rates are not clearly related. As far as I know, no studies clearly showed such a relationship empirically. In arthropods, some taxa that are clearly K-strategist are extremely diverse (by example: the ants), and there are many other factors that can drive speciation rates (for example, karyotype evolution or dispersion abilities). I also think that more generally, linking genetic diversity directly to the ecological ability to occupy diverse niche is not necessarily true: most of the genomic variation is purely or nearly neutral regarding natural selection, species with high genetic diversity are not necessarily more generalist than the others.

>6 (Group 1D). There are models that estimate effective population size accounting for asymmetrical sex ratios, fluctuation in population size. Do you think it would be possible to create a mathematical model to estimate Ne that considers life-history characteristics such as those used in your study, including reproductive strategies?

I think it would be difficult because life-history characteristics should be more considered as Ne proxies rather than clear parameters to measure Ne mathematically. We know mathematically how asymmetrical sex ratios or fluctuation in population size can affect Ne, but the only thing that we know about life-history traits is that they can summarize well the Ne, but not exactly how.

>7. (Group 6D). In their work, ÒEffects of life history traits on genetic diversity in plant speciesÓ, Hamrick and Godt (http://rstb.royalsocietypublishing.org/content/351/1345/1291.short) reported a correlation between mating system and allozyme diversity in plants. Considering that plants may also follow the classification of r and K strategists, do you think it would be likely that a similar approach would reproduce your results?

Yes, I do think that a similar approach in plants would reproduce our results, and I can even give you a link to a recent study that proves it directly: https://academic.oup.com/mbe/article/34/6/1417/3049540 .

>8 (Group 7D). This question is about the model of how r/K strategies influence genetic diversity, assuming that there are stochastic fluctuations in population size (Supplementary Information). Is it reasonable to ascribe the difference between r/K strategists to the Allee effect, since there are some ÒrÓ strategists that donÕt seem to depend on the population density as heavily as others? To try to disentangle the contribution of the Allee effect and the life history traits to the determination of genetic diversity, would it be possible to identify species which share a similar position in the r/k spectrum but differ in having or not the Allee effect?

Yes, it's reasonable to think that the Allee effect might be an important factor here. We though specifically to this at some point and tried to disentagle the effect by some life-history traits, but we did not find obvious metrics easily avalable to measure the dependance to density. For example, we tried to use the dispersion ability of a species (with the idea that individuals from species that can move easily can easily reach sexual partners and are less dependant to density) but it's extremely difficult to homogenize this metric among the large diversity of species that we had. For example, mussels can't move, but they practice external fecundation, so should we use the dispersion ability of the adult or the eggs that they release in the sea? What is the dispersion ability of a mussel egg? Earthworms can move, but I guess that they are more dependent to density than mussels. Also, population structure should have a role, I guess that marine habitats (where all population are more or less interconnected) are less dependent to density than terrestrial (particularly insular) habitats. I think that it would be an interesting question, but I would definitely use a more specific and taxonomically more restricted dataset of species to answer to this question.


Best regards,

Jonathan

Última atualização: quarta-feira, 25 abr. 2018, 17:49