Category Archives: Interviews

Interview with the author: Understanding natural selection and similarity: Convergent, parallel and repeated evolution

Posted on by
Reply

I recently had a chat with José Cerca about his recent piece in Molecular Ecology about how we define parallel and convergent evolution. Questions are in bold.

What got you interested in this topic?

I was always fascinated by how the environments drive evolutionary change, and parallel/convergent evolution are really fascinating as they allow us noticing natural selection and seeing what just keeps evolving across the tree of life.

Great! This paper obviously stems from getting tangled on the various definitions of parallel v. convergent evolution. Was there a project in particular that got you thinking along these lines?

Yeah! So this goes back to another MolEcol paper, which I co-first-authored with Will Sowersby ( https://onlinelibrary.wiley.com/doi/abs/10.1111/mec.16139). Will and his team were focusing on the midas’ cichlid fishes, where these thick lips (Amphilophus labiatus) were present in two different lakes. Will is an ecologist, and his data clearly showed that the A. labiatus had significantly different lips and diet than other species in the lake, A. citrinelus. When I did the genomics of the system … it was quite a mess to untangle. The PCA showed some geographical segregation of PC1 and a weak’ish species segregation on PC2 – so both A. citrinelus and A. labiatus were segregating on PC2 let’s say, ‘in parallel’. This was a mess and we had hybridization across the board, and a lot of phylogenetic mess. So when it was time to write the paper, we entered this rabbit hole: is it parallel? Is it not? In the end we said it wasn’t. A collaborator, David Marques, wrote to me personally saying that, for him, that was clearly parallel evolution … and this made me want to read up the literature and find a good answer (spoilers – I ended up agreeing with David!)

That’s pretty relatable! That’s an anxiety that we all share, right – that some one we look up to will point out a flaw in our work after publication. Even though it is nerve wracking, it is clearly constructive – it certainly has been in this case!

Yeah! Regarding flaws … One thing I liked about the ‘eLife experiment’ is seeing papers as ‘not writen in stone’. Papers can have mistakes, as long they’re not ‘unscientific’.… But ultimately, it’s always nerve wracking.

The paper largely deals with semantics, which terms to use and when. Do you think semantics are particularly important?

Yes. Totally. I think semantics are important; two reasons: first, as we diversity science. As a non-native speaker I regularly have to be constantly checking nuances on terms. Second, I think many times we could avoid unproductive discussions if we agreed our definitions are just not the same. Semantics is this thing that allows us to see whether we disagree in the mechanisms/interpretations, or just in the way we define things. I learnt this in a debating competition: sometimes just nailing down the definition of a term gets us really much further. One word I think a lot about semantics is respect: does ‘I respect that!’ mean you tolerate something, or that you look up to that? I am not really certain and I am sure if you ask different people from different social backgrounds/countries/first-languages, you’ll get different answers.

I agree with you – even if sometimes arguments around semantics can be quite draining! The argument around parallel/convergence probably gets repeated fairly often.

I really hope the paper makes a strong argument to stopping these discussions and that we can all get behind repeated evolution. As I went through different cases I kept thinking: it doesn’t matter whether we call this parallel/convergent evolution, as long as we recognize it repeatedly happens. An alternative – not explored in the paper, and which is messier – is encouraging people to state their definition of parallel/convergent on their paper introductions. But again, this will keep the mess.

I kept getting back to the cichlid work of Claudius Kratochwil. He has these elegant (natural) experiments on transposable element (TE) evolution in cichlids and I kept thinking … it’s kinda cool things just re-evolve and how his work shows the genomic basis of all these things, but perhaps we do not even care for the initial phenotype. For instance, the evolution of melanic stripes in fishes – does it really matter if before selection there were circles, or no melanin at all? Perhaps that may be cool for someone interested in phenotypes, but if your question is on molecular mechanisms – not so much; and so, you can just call it repeated evolution of stripes and move on.

Yeah, I think that would get messy! In the paper, you talk about “gene reuse” as opposed to parallel/convergent evolution. So that’s a kind of catch all. Do you think it matters that population genetic studies may not be looking at phenotypes at all?

I just really think that things get really messy at the genetic level. I kept going through this examples that are on the paper, say – if a phenotype re-evolves from different paralogs, that are 99.9% similar. Is it parallel? Is it convergent? It’s just a rabbit hole. Where do you draw the line? In the end, and mostly due to the extremely helpful review of Maddie James and editorial work of Loren Rieseberg, I just went and got gene reuse, and went for: let’s just circumscribe parallel/convergent evolution to the phenotype. I really think it became a more robust framework because of that – but I also expect that to ruffle the feathers of molecularly oriented people 🙂

I got the gene reuse umbrella from a really cool paper from some colleagues of yours in BC (https://royalsocietypublishing.org/doi/full/10.1098/rspb.2012.2146) [TB is based at UBC in Vancouver]

Right, but what if you looked at some signal of selection (e.g. sweep signals or something like that) and find patterns in homologous regions. At the phenotype perhaps there have been anti-collinear trajectories, to use your term. Does that matter?

I think the terms collinear/confluent are not mutually exclusive with convergent/parallel evolution – they just explore different dimensions: the first on how phenotypes are evolving and the latter about observed/end points.
If you find signal of selections, and experimentally validate them (we don’t do this enough in evolutionary biology, given how convenient genomics became), and find patterns in homologous genes, you can just say – hey I have parallel evolution (if you show ancestral phenotypes to be the same) with gene reuse at homologous genes. Does it make sense?

Yes, I think so. Ok, last question: What was the hardest part about this project?

There were two particular aspects. First, the semantic, because I noticed sometimes I’d send the manuscript to people and they’d be put off by statements which to me were very reasonable. When I asked them about what they meant and we sat and chatted – it ended up being about semantics (of simple things – such as similarity). Second, it was putting the framework together. If you see the acknowledgements (big shoutout to everyone there), people kept coming up with holes and with exceptions which sometimes made me have to revamp the whole thing.

Cerca, J. (2023). Understanding natural selection and similarity: Convergent, parallel and repeated evolution. Molecular Ecology. https://doi.org/10.1111/mec.17132

Interview with the authors: Drivers of genomic landscapes of differentiation across a Populus divergence gradient

Posted on by
Reply

In a recent paper in Molecular Ecology, Shang et al. used population genomic data to characterize patterns of genetic diversity and divergence across the genomes of 8 different species of poplar. With an extensive population genomic dataset, they characterize patterns of introgression and signals of linked selection across the various species’ genomes. They found striking correlations in the landscapes of genetic variation across the species’ genomes, suggesting conserved patterns of linked selection across the divergence gradient of their sampled species.

We sent some questions to Huiying Shang and Thibault Leroy, the corresponding authors of this work, to get more detail on this study.

 Populus tomentosa. Photo credit: Yuan Lu

What led to your interest in this topic / what was the motivation for this study? 

We are fascinated by speciation, the process that generates biodiversity. An ongoing challenge in the discipline is  identifying so-called speciation genes, i.e., molecular variation that generates reproductive barriers to gene flow. One general strategy to identify these genomic regions involves scanning across pairs of genomes and looking for specific signals in variation of nucleotide diversity and differentiation across the genome. However, similar signals amongst the genomic “landscapes” can be generated by other evolutionary forces, in such a way that the landscapes are known to be highly heterogeneous and therefore complex to interpret. Relatively little remains known about the relative contribution of the different evolutionary forces and how these landscapes evolve with time. Fascinated by some pioneering works in monkeyflowers (Stankowski, et al. 2019 PLoS Biology 17) and in avian species (Irwin, et al. 2018 Molecular Ecology 27), Christian Lexer planned to use a series of Populus species pairs along a divergence gradient as models, with pairs spread from an early to a late stage of speciation.  In addition, the genus Populus has a wide geographic distribution, a small diploid genome,  and experiences extensive gene flow between some species pairs. We therefore decided to resequence 200 whole-genomes to investigate the relative contribution of different evolutionary processes to these landscapes.

What difficulties did you run into along the way? 

Our general hypothesis was the main driving force promoting genetic differentiation changes during the course of speciation. Reflected in the genome, different drivers lead to differentiated genomic footprints regarding the local variation in nucleotide diversity (π), relative (FST) and absolute sequence divergence (DXY). By investigating the genomic landscapes of a series of species pairs along a divergence gradient, we thought we can  identify the main evolutionary force that contributes to the genomic landscapes of differentiation based on theoretical expectations under different evolutionary scenarios. But unfortunately, it is not as simple, in fact, linked selection played a prime role in shaping the genomic landscapes across all species pairs. Christian Lexer was the main initiator of many of these ideas for the project, the sudden loss of our colleague drove us all to see this project through in honour of his memory. We think he would have been fascinated with the scientific journey this project ultimately provided and the new opportunities opening up in speciation genomics

What is the biggest or most surprising innovation highlighted in this study? 

We found significantly positive correlations between DXY and π in all species pairs, though the correlation coefficients became weaker as divergence increased. This was not expected, as in the early stage of speciation, we observed extensive gene flow between species and expected our results to be in agreement with a ‘divergence with gene flow’ scenario (including negative correlations between DXY and π). Even at early stages of divergence,  our results were more consistent with a prime role of linked selection, especially background selection, in shaping the genomic landscapes.

Moving forward, what are the next steps in this area of research?

Our approach is interesting but remains limited by its empirical nature, because our interpretation is mostly based on verbal models. Methodological developments allowing new methods to compare the different scenarios – virtually allowing to quantify the associated parameters – could be extremely helpful in order to aid future interpretations.  In the future, the development of such methods, trained with extensive demo-genetic simulations, could greatly contribute to a better understanding of the effects of the divergence process on genome-wide molecular patterns.

Describe the significance of this research for the general scientific community in one sentence.

Speciation is a complex continuous process involving multiple evolutionary factors contributing to genomic differentiation, but the main force involved may vary at different stages.

Describe the significance of this research for your scientific community in one sentence.

We found that the genomic landscapes of diversity and divergence are mostly shaped by linked selection, along with gene flow and standing genetic variation, and that this result holds true independently of the stage across the divergence gradient.

Shang, H., Field, D. L., Paun, O., Rendón-Anaya, M., Hess, J., Vogl, C., Liu, J., Ingvarsson, P. K., Lexer, C., & Leroy, T. (2023). Drivers of genomic landscapes of differentiation across a Populus divergence gradient. Molecular Ecology, https://doi.org/10.1111/mec.17034

Interview with the authors: Landscape heterogeneity and Quaternary climate fluctuations have shaped the demography and present-day distribution of genetic variation of a tropical oak

Posted on by
Reply

In a recent paper in Molecular Ecology, Ortego et al. integrated genomic data of the Andean oak  (Quercus humboldtii) from the Colombian Cordilleras and spatiotemporally explicit inference frameworks to test competing hypotheses about the demographic trajectories of the species. Their analyses support a dynamic model of flickering population connectivity since the Last Glacial Maximum, emphasizing the interplay between spatial configuration of geographical barriers to dispersal and Quaternary climatic oscillations in shaping demographic trajectories of montane oak forests, and further, tropical montane biotas.

We sent a number of questions to lead author of this work, Joaquín Ortego, to get more detail on this study.

Forest of the tropical oak Quercus humboldtii in the Colombian Andes. Photo credit: Josep Maria Espelta

What led to your interest in this topic / what was the motivation for this study? 

Relatively little is known about the impacts of Pleistocene glacial-interglacial cycles on tropical biotas and the different models proposed to explain their hypothetical demographic and distributional shifts in response to past climate changes have seldom been formally tested using genomic data. Thus, the main motivation of our study was to shed some light into the processes that have determined the demographic dynamics of Andean montane forests, with a particular emphasis on the potential role of Quaternary climatic oscillations on shaping population connectivity and isolation. To set up our hypotheses, we found very inspiring the extensive palaeoecological research performed in the tropical Andes during the past decades (e.g., Hooghiemstra & Van der Hammen, 2004) and the alternative distributional and demographic models proposed in previous studies (Ramírez-Barahona & Eguiarte, 2013; Flantua et al., 2019). By integrating genomic data, species distribution modelling and demographic simulations we could explicitly test hypotheses that have remained largely unanswered for a long time.

What difficulties did you run into along the way? 

From an analytical point of view, the main difficulty we found was creating the different spatially explicit scenarios we aimed to test, particularly the one incorporating reconstructions of past species distribution at a fine temporal resolution. It would have been also nice to analyze populations across the entire distribution of the species, but it was not possible due to logistic limitations and time-constrains during our fieldwork.

The three spatially explicit models tested in the study. Genomic data support that Andean oak forests follow a model of flickering genetic connectivity linked to Quaternary climatic oscillations (bottom panel, in green)

What is the biggest or most surprising innovation highlighted in this study? 

The most interesting aspect of our study is that we demonstrate that the Andean montane oak forest is not a static ecosystem. On the contrary, our genomic data and demographic modelling indicate that Andean montane forests have experienced changes in population connectivity and isolation through time linked to Quaternary climatic oscillations. It is also surprising the high genetic fragmentation of tropical oak populations, which has likely resulted in most of them have experienced idiosyncratic demographic trajectories. A totally unexpected result we obtained – actually inspired by the comments of one of the Reviewers – was that genetic differentiation was best explained by elevational dissimilarity among populations (i.e., “isolation-by-elevation”) than by the geographical distance that separate them (i.e., “isolation-by-distance”), pointing to local adaptation processes linked to the contrasting environmental conditions prevailing at different elevational ranges. This result is pretty cool in my opinion, as such a clear effect of elevation dissimilarity on patterns of gene flow among populations has been very rarely reported in the literature. Definitively, I would like to explore this aspect more in depth in the future, analyzing more populations and testing for genomic signatures of local adaptation or phenological mismatching linked to environmental dissimilarity.

Moving forward, what are the next steps in this area of research?

From my perspective, the next natural step would be to extend our analyses to multiple species of the different vegetation belts (i.e., lowlands, lower montane forest, upper montane forest, and paramos) within a community-level comparative framework. This would allow testing for concordant vs. species-specific demographic trajectories across multiple taxa, which might help the reach more general conclusions about the role of past climatic oscillations on the dynamics of tropical ecosystems (e.g., Helmstetter et al., 2020; Prates et al., 2016). The aforementioned testing of “isolation-by-elevation” in other taxa might also provide interesting insights into the proximate ecological and evolutionary processes that have contributed to geographical diversification in the tropical Andes and the extraordinary levels of local endemism characterizing this biodiversity hotspot.

Detail of leaves and acorns of the tropical oak Quercus humboldtii. Photo credit: Josep Maria Espelta

What would your message be for students about to start developing or using novel techniques in Molecular Ecology?

My message would be that they do not get lost in the massive genomic data that we can now easily obtain and focus on the questions and hypotheses that they want to address. Curiosity moves science forward, simple observation and exploration is the beginning of everything!

What have you learned about methods and resource development over the course of this project?

The 100-year resolution bioclimatic layers now available through the CHELSA-TraCE21k database (https://chelsa-climate.org/chelsa-trace21k/; Karger et al., 2023) allowed us to get a more realistic picture of the distributional dynamics of the focal species, which was key for our demographic inferences and interpreting the data (e.g., changes in elevational displacements through time). I think this new resource will be very useful for future biogeographic and distributional studies.

Describe the significance of this research for the general scientific community in one sentence.

Tropical ecosystems were not spared from the impacts of Pleistocene climatic changes.

Describe the significance of this research for your scientific community in one sentence.

Andean montane forest follow a model of flickering genetic connectivity linked to Quaternary climatic oscillations.

References

Flantua, S. G. A., O’Dea, A., Onstein, R. E., Giraldo, C., & Hooghiemstra, H. (2019). The flickering connectivity system of the north Andean paramos. Journal of Biogeography, 46(8), 1808–1825. https://doi.org/10.1111/jbi.13607

Helmstetter, A. J., Bethune, K., Kamdem, N. G., Sonke, B., & Couvreur, T. L. P. (2020). Individualistic evolutionary responses of central African rain forest plants to Pleistocene climatic fluctuations. Proceedings of the National Academy of Sciences of the United States of America, 117(51), 32509–32518. https://doi.org/10.1073/pnas.2001018117

Hooghiemstra, H., & Van der Hammen, T. (2004). Quaternary ice-age dynamics in the Colombian Andes: Developing an understanding of our legacy. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 359(1442), 173–180. https://doi.org/10.1098/rstb.2003.1420

Karger, D. N., Nobis, M. P., Normand, S., Graham, C. H., & Zimmermann, N. E. (2023). CHELSA-TraCE21k – High-resolution (1 km) downscaled transient temperature and precipitation data since the last glacial maximum. Climate of the Past, 19(2), 439–456. https://doi.org/10.5194/cp-19-439-2023

Prates, I., Xue, A. T., Brown, J. L., Alvarado-Serrano, D. F., Rodrigues, M. T., Hickerson, M. J., & Carnaval, A. C. (2016). Inferring responses to climate dynamics from historical demography in neotropical forest lizards. Proceedings of the National Academy of Sciences of the United States of America, 113(29), 7978–7985. https://doi.org/10.1073/pnas.1601063113

Ramírez-Barahona, S., & Eguiarte, L. E. (2013). The role of glacial cycles in promoting genetic diversity in the neotropics: The case of cloud forests during the last glacial maximum. Ecology and Evolution, 3(3), 725–738. https://doi.org/10.1002/ece3.483

Featured study

Ortego, J., Espelta, J. M., Armenteras, D., Díez, M. C., Muñoz, A., and Bonal, R. (2023) Demographic and spatially-explicit landscape genomic analyses in a tropical oak reveal the impacts of late Quaternary climate change on Andean montane forests. Molecular Ecology, 32(12), 3182-3199. https://doi.org/10.1111/mec.16356

Interview with the authors: Unparallel differentially expressed genes in parallel ecological divergence

Posted on by
Reply

In a recent paper in Molecular Ecology, Szukala et al. quantified the degree of gene expression and functional parallelism across polytopic divergence of montane and alpine ecotypes of in Heliosperma pusillum (Caryophyllaceae) and gained insights into the architecture of adaptive traits. They performed RNA-seq analyses of plants grown in a common garden and detected a large proportion of differentially expressed genes in each replicate ecotype pair. Functional enrichment of these genes, however, revealed that the traits affected by significant expression divergence are largely consistent across ecotype pairs, suggesting a polygenic architecture for the diverged adaptive traits and multiple routes for adaptation. A new genome assembly for H. pusillum was also presented in this study.

We sent a number of questions to lead authors of this work, Aglaia Szukala and Ovidiu Paun, to get more detail on this study.

Upper panels: Graphical representations of the alpine (A) and montane (M) ecotypes of Heliosperma pusillum and their ecological niches. Lower panel: Map showing the study sites of ecotype pairs that evolved in parallel.

What led to your interest in this topic / what was the motivation for this study? 

We are fascinated by the concept of parallel evolution and the molecular mechanisms behind this process. Given that drift is a major driver of evolution and due to the traditional focus on mono- or oligogenic traits, parallel evolution has been considered to be a rare process until recently. However, together with the increasing understanding of polygenic adaptation (Barghi et al., 2020), it has become clear that parallel evolution is relatively frequent, with implications across evolutionary biology and ecology. Specifically, for our study, previous works (Trucchi et al., 2017; Bertel et al., 2018) reported some evidence that altitudinal ecotypes in H. pusillum diverged in parallel. We wanted to rigorously test this hypothesis using demographic modeling and understand the level of molecular parallelism, with regard to divergent gene expression and outlier SNPs.

What difficulties did you run into along the way? 

For several reasons related to the planning of field work and the development of the wider project over years, we had to deal with uneven sampling sizes across populations, which needed to be taken into account, especially for the demographic modeling analyses. Fun fact: in reciprocal translocation experiments of a complementary study (Szukala et al., 2022) on the same species, whose data is also included in the present paper, we chose to use alpine microsites to plant our accessions that were fairly flat (in an otherwise steep area) and free of other plants. At the end of the vegetation season, those sites proved to be resting places for chamois which squeezed and munched most of our plants, while overfertilizing them. In previous years, when reciprocal transplantations were performed as preparation for this study, we faced droughts, poor germination and survival rates at some sites, leading to uneven sampling sizes across sites. Take-home message: experiments in the wild are always a challenge.

What is the biggest or most surprising innovation highlighted in this study? 

It is unclear how much overlap of divergence outliers is to be expected across natural evolutionary replicates. Our study showed a surprisingly low amount of shared molecular differentiation, which we did not expect given that the geographic range considered is relatively small and our study system is in a phase of incipient speciation with no reproductive isolation detectable (Bertel et al., 2016). The extremely low sharing of differentially expressed genes and outlier SNPs, but high similarity of GO terms involved across independent divergence events, indicates that the polygenic architecture of traits is relevant for adaptation of these populations to distinct altitudinal zones in the Alps.

Moving forward, what are the next steps in this area of research?

We further investigated the role of phenotypic plasticity for the development of parallel evolution in our system (Szukala et al., 2022) towards a better understanding of the relative role of genetic and environmentally-induced phenotypic variation in such replicated divergence events. We are also interested if other molecular mechanisms, which are sensitive to environmental input, such as epigenetic signals, could play an important role in parallel evolution. Further, we wish to understand how polygenic adaptation affects signatures of parallel evolution. Very interesting is to question if adaptation can use different genes to produce similar outcomes even in very closely related lineages, and how frequent this process takes place compared to the re-use of standing variation.

Alpine (blue) and montane (light orange) ecotypes of Heliosperma pusillum and their environments. In the alpine environment glabrous plants grow on more humid screes and meadows, also in proximity of streams. In the montane environment below the tree line, pubescent plants typically grow under rocks overhangs or on the rock as chasmophytes. Photo credit: Szukala A and Paun O.

What would your message be for students about to start developing or using novel techniques in Molecular Ecology?

Being curious and exploiting the most advanced and newest methods is good, but don´t forget to be robust, careful, and bias-aware when it comes to the interpretation of results.

What have you learned about methods and resource development over the course of this project?

It is often difficult to quantify and describe results relative to expectations in an objective way, because it is hard to formulate objective expectations in natural systems.

Describe the significance of this research for the general scientific community in one sentence.

Repeated evolution of similar phenotypes can involve different sets of genes.

Describe the significance of this research for your scientific community in one sentence.

Polygenic traits offer different genetic substrates for parallel evolution of similar phenotypes.

References

Barghi N, Hermisson J, Schlötterer C. 2020. Polygenic adaptation: a unifying framework to understand positive selection. Nature Reviews Genetics 21: 769–781.

Bertel C, Hülber K, Frajman B, Schönswetter P. 2016. No evidence of intrinsic reproductive isolation between two reciprocally non-monophyletic, ecologically differentiated mountain plants at an early stage of speciation. Evolutionary Ecology 30: 1031–1042.

Bertel C, Rešetnik I, Frajman B, Erschbamer B, Hülber K, Schönswetter P. 2018. Natural selection drives parallel divergence in the mountain plant Heliosperma pusillum s.l. Oikos 127: 1355–1367.

Trucchi E, Frajman B, Haverkamp THA, Schönswetter P, Paun O. 2017. Genomic analyses suggest parallel ecological divergence in Heliosperma pusillum (Caryophyllaceae). New Phytologist 216: 267–278.

Szukala A, Bertel C, Frajman B, Schönswetter P, Paun O. 2022. Parallel adaptation to lower altitudes is associated with enhanced plasticity in Heliosperma pusillum (Caryophyllaceae). bioRxiv 2022.05.28.493825; doi: 10.1101/2022.05.28.493825.

Featured study

Szukala A, Lovegrove-Walsh J, Luqman H, Fior S, Wolfe TM, Frajman B, Schönswetter P, Paun O. 2022. Polygenic routes lead to parallel altitudinal adaptation in Heliosperma pusillum (Caryophyllaceae). Molecular Ecology. https://doi.org/10.1111/mec.16356.

Interview with the authors: Genomic basis of Y‐linked dwarfism in cichlids pursuing alternative reproductive tactics

Posted on by
Reply

In a recent paper in Molecular Ecology, Singh et al. used genome sequencing, bioinformatics and population genetic analyses to gain insight into the genetics and evolution of a fascinating mating system. The species in question, Lamprologous callipterus, exhibits a mating system with two males morphs. Large “bourgeois” males carry empty snail shells that are inhabited and used as nests by the females. An alternate male morph, much smaller than the “bourgeois” males, also exists and inhabits shells along with the females. Previous genetic work had established that this mating system was Y-linked and that the male body size was a Mendelian trait, but the sex-determining locus had not been identified until this study.

We sent some questions to Pooja Singh, the author who led this work, to get more detail on this study.

Photo credit: Drawing by Pooja Singh, based on Barbara Taborsky’s original image.

What is the biggest or most surprising innovation highlighted in this study? 

The most novel aspect of this study is that we found an example of a young sex chromosome that may have evolved due to sexual antagonism over body size. While the sexual antagonism theory is considered the classical model of sex chromosome evolution, few empirical examples exist to support it. The other exciting finding was that the candidate body size/dwarfism gene that we propose for L. callipterus, GHRHR, is a well-known dwarfism gene in mammals. Fish and mammals shared a common ancestor over 440 million years ago, so the body size development pathway is genetically constrained through deep evolutionary time.

What difficulties did you run into along the way? 

The major challenge for me was that I knew little about sex chromosome evolution when I started this project, so I really had to do a lot of groundwork reading relevant literature and researching methods to be able to get things going. I had to start thinking beyond the classical XY and WZ old sex systems and familiarize myself with the workings of early stages of sex chromosome evolution.

When I read your paper, I had never heard of the fascinating mating system of these cichlids. They reminded me of the ruff, and the multiple inversions that seem to be involved in the different reproductive strategies in that system. You mention in the paper that you were not able to identify inversions based on the bioinformatic approaches you used. Is there a sense for how much chromosome evolution during the radiation? Could the use of the divergent reference genome have anything to do with the lack of a signal of inversions? 

To my knowledge the broad scale chromosomal structure of African cichlid species is similar. However, small scale structural variations (inversions, indels, translocations etc.) have not been investigated systematically. So yes, it is totally possible that our short read data and the divergent (and fragmented) reference genome assembly may have hindered our ability to detect inversions. The system really needs a long-read de novo genome assembly to resolve the inversion question.

In the Discussion, you talk about the possibility of different male Y-haplotypes. Is your data sufficiently high resolution that you could examine insertion/deletion polymorphisms in your dataset? 

Yes, we could technically identify small insertions/deletions in our data. Might certainly be something to investigate in the future, in combination with long-read Y assembly.

A recently proposed model of sex-chromosome evolution indicates that gene expression differences may predominate at the early stages of sex chromosome evolution (Lenormand and Roze 2022 Science – https://doi.org/10.1126/science.abj1813). This is intriguing given that you didn’t find any smoking gun loci with signals of sexual antagonism. Do you have plans to look at patterns of gene expression across the different morphs?

While Lenormand and Roze’s theory is certainly exciting for the field of sex chromosome evolution, I think it is less plausible for the L. callipterus mating system because antagonistic body sizes in females and males are crucial shell-brooding success and fitness. And because we found the candidate sex-determining gene and body-size gene to be physically linked. I am certainly not a proponent of the ‘one classical theory explains all’ narrative and I really look forward to seeing what RNA-seq data reveal about sex chromosome evolution in this species. It would be especially interesting to see the landscape of cis/trans eQTLs of genes in our proposed L. callipterus sex chromosome and how much it reflects the expectations from Lenormand and Roze’s model. Beyond just this species, cichlid fishes are an excellent system to test the sexual antagonism vs gene regulation models of sex chromosome evolution.

Regarding the coverage analysis you used to identify the putative sex-linked locus. Given the hypothesis that the divergence of the sex chromosomes is recent, reads sampled from Y-linked regions may still map well to the orthologous region on the putative X-chromosome.  Did you tweak mapping quality filters at all?

I did run a less stringent mapping analysis, which overall had slightly higher mapping statistics, but the reduced coverage pattern on the L. callipterus sex chromosome was still significant.

Snail shell nest of L. callipterus with the nest owner in the left centre.
Photo credit: Koblmueller et al. 2007

Could you describe the significance of this research for the general scientific community in one sentence?

Sexual antagonism over body-size may have driven sex chromosome evolution in a shell-brooding cichlid fish where giant and dwarf male reproductive types have evolved.

Moving forward, what are the next steps in this area of research (unless otherwise covered)?

Our main priority right now is to keep the L. callipterus dwarf males alive and breeding. Our fish were recently moved from the University of Bern in Switzerland to the University of Graz in Austria, and it has proved difficult to get the dwarf males happy and breeding in the new facility. This is (probably) the only collection in the world of L. callipterus dwarf males outside Lake Tanganyika so they are very precious. Our next step is to write a convincing grant to get funding to build a long-read improved genome assembly, conduct RNA-seq, and sample and sequence more individuals from natural populations. I would like to use the RNA-seq data to map expression QTLs and investigate the regulatory interactions of candidate genes related to sex, morphology, physiology, and behaviour that we found in or around the L. callipterus sex region. It would also be interesting to study sex chromosomes in related lamprologine species, as our pre-liminary analysis in this manuscript suggests that the sex region may be shared across the lamprologine tribe

Interview with the authors: Whole-genome analysis of multiple wood ant population pairs supports similar speciation histories, but different degrees of gene flow, across their European ranges

Posted on by
Reply

In a recent paper in Molecular Ecology, Portinha et al. used population genomic data to analyse the speciation history of two closely related species of wood ants, Formica polyctena and F. aquilonia. Using a demographic modelling approach, the authors reconstruct the history of divergence for multiple heterospecific pairs of populations. In all cases, the authors found that there was evidence for divergence with gene flow. However, for a sympatric population pair sampled in Finland there was evidence for substantially elevated gene flow between the species. Their findings imply that population genomic analysis of speciation history may be geographically variable for particular species.

We sent some questions to Beatriz Portinha and Pierre Nouhaud, the corresponding authors of this work, to get more detail on this study.

Ant mound surface covered in ants. Photo credit: Jack Beresford

What led to your interest in this topic / what was the motivation for this study? 

Knowledge on the demographic and speciation histories is essential for understanding
contemporary genomic patterns in natural populations, which is why we wanted to
reconstruct it for the emerging Formica model system. Our study species, Formica polyctena
and F. aquilonia, are known to hybridize naturally in Southern Finland, where their hybrids
have been studied for over 10 years (Kulmuni et al., 2010; Martin-Roy et al. 2021). We
wanted to test whether a similar divergence history was consistently inferred across the
European ranges of both species, or whether the Finnish populations would stand apart,
possibly because of gene flow mediated by hybrid populations in the area.

What difficulties did you run into along the way? 

Formica polyctena and F. aquilonia had a limited genomic toolbox when we started the
project, and we initially relied on a distant and non-contiguous reference assembly.
Meanwhile, our group assembled a high quality reference genome (Nouhaud et al., 2022),
which improved the quality of our inferences.


The demographic modelling software we used, fastsimcoal2, can simulate a large panel of
evolutionary scenarios. When planning this study, we wanted to design models that
considered alternative scenarios for the divergence of the species which would be as
biologically meaningful as possible, while keeping the number of models low enough that the
project 1) would not be a huge computational burden and 2) would be executable in the
available time frame (Beatriz’s MSc. project, funded by Erasmus+ and Societas pro Fauna et
Flora Fennica). This was an especially important aspect as we used four distinct population
pairs to reconstruct the history of the two species, so each model had to be run, at least, four
different times.

What is the biggest or most surprising innovation highlighted in this study? 

We found that there was already bidirectional gene flow occurring in Finland before the
hybridization events that led to the present-day hybrid populations. This was not suspected
before, as there is no evidence in the literature, and it suggests that F. polyctena in Finland
may be admixed, which is supported by the fact that we have not found non-admixed F.
polyctena individuals in Finland.

Moving forward, what are the next steps in this area of research?

The divergence history we inferred between F. polyctena and F. aquilonia can be used to
run simulations about the evolution of the hybrid populations, which is what we did in a
subsequent work (Nouhaud et al. 2022). In the longer run, it would also be important to
extend this work by reconstructing the divergence history of the whole F. rufa species group,
which encompasses 5 species (including F. aquilonia and F. polyctena) and where gene flow
is prevalent (Seifert, 2021).

Describe the significance of this research for the general scientific community in one sentence.

Genomes from individuals sampled thousands of kilometers apart tell the same ancient
history, while their most recent history may be different.

Describe the significance of this research for your scientific community in one sentence.

The divergence history between two species can be reliably and consistently inferred from a
small number of individuals sampled across the species’ ranges.

Portinha, B., Avril, A., Bernasconi, C., Helanterä, H., Monaghan, J., Seifert, B., Sousa, V. C., Kulmuni, J., & Nouhaud, P. (2022). Whole-genome analysis of multiple wood ant population pairs supports similar speciation histories, but different degrees of gene flow, across their European ranges. Molecular Ecology, 31, 3416– 3431.

Kulmuni, J., Seifert, B. & Pamilo, P. (2010). Segregation distortion causes large-scale
differences between male and female genomes in hybrid ants. Proceedings on the National
Academy of Sciences, 107(16), 7371-7376.


Martin-Roy, R., Nygård, E., Nouhaud, P. & Kulmuni, J. (2021). Differences in thermal
tolerance between parental species could fuel thermal adaptation in hybrid wood ants.
American Naturalist, 198(2), 278-294.


Nouhaud, P., Beresford, J. & Kulmuni, J. (2022). Assembly of a hybrid Formica aquilonia× F.
polyctena ant genome from a haploid male. Journal of Heredity, esac019, 1-7.


Nouhaud, P., Martin, S. H., Portinha, B., Sousa, V. C. & Kulmuni, J. (2022). Rapid and
repeatable genome evolution across three hybrid ant populations. bioRxiv.


Seifert, B. (2021). A taxonomic revision of the Palaearctic members of the Formica rufa
group (Hymenoptera: Formicidae) – the famous mound-building red wood ants.
Myrmecological News, 31, 133-179.

Interview with the authors: Mega-fire in redwood tanoak forest reduces bacterial and fungal richness and selects for pyrophilous taxa that are phylogenetically conserved

Posted on by
Reply

In a recent paper in Molecular Ecology, Enright et al. examined how soil microbiomes are affected by extreme fires. The Soberanes mega-fire provided the authors with an opportunity to study how such extreme events, which are increasingly common with climate-change, can have lasting effects on ecology. By sampling the soil microbiome before and after the Soberanes mega-fire, Enright at al. demonstrated dramatically altered soil communities and a reduction in species richness associated with the mega-fire. There was a clear phylogenetic pattern to the particular microbes that increased or decreased abundance after the fire. Drawing from their results, Enright et al. propose a framework to predict the traits that post-fire microbial communities might exhibit.

We sent some questions to Sydney Glassman, one of the corresponding authors of this work, to get more detail on this new study.

Aerial view of the Soberanes mega-fire. Photo credit: Calfire

What led to your interest in this topic / what was the motivation for this study? 

I had originally been interested in sampling the redwood tanoak forests of Big Sur because I was interested in what the cascading effects of sudden oak death (SOD) induced mortality would be on soil fungal communities during my PhD at UC Berkeley. Prof Dave Rizzo at UC Davis had a large plot network investigating the effects of SOD on plant mortality. I teamed up with him in 2011 to select a subset of plots to collect soils to investigate the impacts on the soil microbial community via amplicon sequencing. Then, in 2016, I learned that half my plots burned in the catastrophic Soberanes Megafire. It’s extremely rare to have pre- and post-fire samples from the same sampling locations before and after a mega-fire. I was really curious about what the impact of a mega-fire would be on soil microbial communities especially since they had never been studied in redwood tanoak forests before. These forests are endemic and charismatic megflora of Califronia that are facing multiple global change factors and it is really unclear how the soil microbial communities will respond to wildfires and how that will influence the recovery of the vegetation. I had already moved to southern California at this time to start a post-doc at UC Irvine, so I asked Kerri Frangioso, who lived in Big Sur, if she would be able to re-sample any of the plots that burned. Using GPS, she was able to collect soils from the exact same sampling locations that I had sampled in 2011 from 3 of the plots (2 burned and 1 unburned) within 30 days of the fire being declared over. She mailed these soils to me, I extracted the DNA, and froze everything until I was able to start my own lab at UC Riverside in 2018.

What difficulties did you run into along the way? 

The terrain in Big Sur is notoriously challenging to traverse. It is extremely steep, lots of windy dirt roads, and there is a lot of poison oak. There is no cell reception in any of our plots and most are at least an hour from the nearest town.  Collecting the soil even before the fire was challenging enough. However, after fires, it is really challenging to access sites because roads are closed, landslides are common, and dead or dying trees are extremely hazardous especially in the case of wind. We were very lucky to be able to re-sample even 3 of our plots so fast after the fire.

What is the biggest or most surprising innovation highlighted in this study? 

I was really surprised that many of the same pyrophilous “fire loving” microbes that have been found to increase in frequency after pine forest fires also increased in frequency after redwood tanoak fires. That indicates that soil microbes are selected for by slightly different pressures than plants because the plants that regenerate post-fire in pine forests vs redwood tanoak forests are very different. It seems more likely that microbes instead survive via temperature thresholds and if fire is high severity enough, similar groups of microbes will respond. We collaborated with Kazuo Isobe to implement the CONSENTRAIT analysis and identified that microbial response to fire was indeed phylogenetically conserved, and it seemed that related groups of bacteria and fungi did indeed positively or negatively respond to fires. This will greatly enhance our ability to predict which microbes will respond to fire in any ecosystem since certain lineages seem evolutionarily adapted to survive fires. We also found that a basidiomycete yeast Basidioascus, dominated the fungal sequences at 30 days post-fire, and that had never been found before, probably because most post-fire sampling historically has been based on fruiting bodies.

Morphological diversity of soil microbes. Photo credit: Jenna Maddox

Moving forward, what are the next steps in this area of research?

I was able to leverage some of these results and results from my work sampling wildfires in Southern California chaparral to help me acquire a USDA grant from their Agricultural Microbiomes program (described here). The purpose of this grant is to characterize the traits of pyrophilous microbes and begin to get our knowledge of fire adaptation in microbes to that of plants. We understand a lot of the traits that enable plants to survive wildfires (like thick bark, vegetative resprouting, serotinous cones, etc) but we don’t have similar understanding of those traits in microbes. In order to understand these traits, Dylan Enright has begun performing biophysical trait assays on these microbes to determine their traits based on a large culture collection of pyrophilous microbes that I have been developing since I started my lab in July 2018. Over the last four years, 2 lab managers, one PhD student (Dylan Enright), 13 UCR undergraduates, and one part time laboratory technician have been involved in developing this culture collection of over 400 isolates of bacteria and fungi from burned soils from wildfires. Our goal is to characterize their traits with biophysical assays and eventually with genomics.

Have you gone back (or have you any plans to go back) to sample soils in the post-fire period? How long lasting do you think the effects of fire on microbial communities would be? 

Unfortunately, I have not been able to get this particular project funded (despite several attempts) and everything I did for this paper was completely unfunded. So I have not been able to return to these plots to sample again. I would be interested in returning to them eventually. I would predict the effects of the fire on the microbial communities could last decades if not longer, depending on if the plants themselves have been able to recover. Most of the literature on pyrophilous microbes suggests that high severity fire can have long term impacts on soil microbes that can last at least a decade or more. Given that the richness of both bacteria and fungi was reduced by up to 70% in one of our plots, I would predict it will take a long time to recover.

Describe the significance of this research for the general scientific community in one sentence.

Megafires have long lasting impacts on both plants and soil microbes alike, and it is important to understand the impacts on soil microbes since they drive plant and soil regeneration. 

Describe the significance of this research for your scientific community in one sentence.

The pyrophilous microbes that respond to a mega-fire in redwood tanoak forests are similar to those that respond to high severity wildfires in better studied pine forest systems, and the fact that they are phylogenetically conserved indicates that we will be able to predict what microbes will respond to wildfires in any system. Further, we are beginning to identify conserved trait responses that enable wildfire response that are analogous to plants and will help us bin and better understand fire adaptation traits in microbes.

Enright, D. J., Frangioso, K. M., Isobe, K., Rizzo, D. M., & Glassman, S. I. (2022). Mega-fire in redwood tanoak forest reduces bacterial and fungal richness and selects for pyrophilous taxa that are phylogenetically conserved. Molecular Ecology, 31, 2475– 2493.

Interview with the authors: Genetic data and niche modeling reveal complex interspecific interactions of invasive species with native congeners and help evaluate distribution pattern, range limits and invasion risk of the species

Posted on by
Reply

In a recent paper in Molecular Ecology, Espindola and Vázquez-Domínguez et al. combined comprehensive fieldwork, genetic analyses and a novel niche modeling approach to investigate population genetic patterns, distribution patterns of native and non-native red-eared slider turtle (Trachemys scripta elegans), one of the worst invasive species across the world, and its congeners. They found very little naturally occurring distribution overlap and genetic admixture between red-eared slider and other Trachemys species studied. In addition, they demonstrated that the native Trachemys species in Mexico have distinct climatic niche suitability, which probably prevents the invasion of red-eared slider in the area. However, major niche overlap was found between non-native red-eared slider and native species from different parts of the world, indicating that sites closer to ecological optima of invasive species have higher establishment risk than those closer to the niche-centre of the native species.

We sent a number of questions to lead authors of this work, Sayra Espindola and Ella Vázquez-Domínguez, to get more detail on this study.

What led to your interest in this topic / what was the motivation for this study? 

We have interest in population genetics of invasive species. In addition, Trachemys scripta elegans, one of the World’s 100 worst invasive species, is native to NE North America, and several native congeneric species are naturally distributed along the eastern coast of Mexico, which is an extraordinary scenario to test the effect of congeners on potential invasion patterns and evaluate their climatic and niche differences.

Trachemys spp (T. scripta, T. venusta, T. cataspila, T. taylori) and their distributions along the west coast of USA and Mexico. Trachemys scripta (red dots) in Mexico is non-native. A turtle trapping net is shown. Figure credit: Sayra Espindola/Ella Vázquez-Domínguez

What difficulties did you run into along the way? 

Maybe the most significant was that, at the time we did the molecular laboratory work, extracting DNA from samples that had been stored in formaldehyde (museum samples) was rather difficult, thus we could not obtain genomic data (SNPs) for those samples (extraction kits are much more efficient now). Nonetheless, we did sequence nuclear microsatellites loci, which provided adequate genetic information that enabled us to show the significant contemporary genetic differentiation present between native and non-native Trachemys scripta elegans individuals. 

What is the biggest or most surprising innovation highlighted in this study? 

There are two interesting findings. One is that non-native Trachemys scripta elegans individuals have very little naturally occurring distribution overlap and admixture with its congeners – they exhibited reduced gene flow and clear genetic separation despite having zones of contact. Also, we demonstrate that the native Trachemys species studied (T. cataspila, T. venusta) have distinct climatic niche suitability, which prevents the establishment of and displacement by the non-native Trachemys scripta elegans. Yet, as T. s. elegans has invaded and displaced native turtle species worldwide, we show that sites closer to T. s. elegans’ niche-center have higher establishment risk than those closer to the niche-center of the native species.

Moving forward, what are the next steps in this area of research?

We are working with our genomic data to identify loci under selection to evaluate the potential connection between specific genes and adaptive traits in these turtles. Considering the distinct climatic niches and distribution we found for the turtles, we are very kin to elucidate if there are adaptive differences among them. In addition, our results set the basis for future work – whole genome or gene-targeted sequencing, as well as a higher number of field-sampled individuals, would allow assessment of hybridization and specific gene introgression.

What would your message be for students about to start developing or using novel techniques in Molecular Ecology?

We would first tell them that molecular ecology research, combining ecological fieldwork and laboratory tasks, is absolutely amazing! We recommend choosing to work with the species/taxa that you more deeply like – this makes the journey very enjoyable; and also selecting a laboratory and research group with ample experience in molecular work and analyses, while at the same time not afraid of proposing novel questions and ways of analyzing them.

What have you learned about methods and resource development over the course of this project?

In this project, we proposed and developed a novel modeling approach, in which by contrasting the niche suitability of the species, we were able to include, indirectly, the interactions that can occur when a species is introduced to habitats occupied by other species. The model is based on analyses of climatic niche suitability and the environmental centrality hypothesis, where fitness is expected to be highest in sites with environments closest to the center of the niche of the species. The development of this model and algorithms required an immense number of trials and errors, and once we had the final version, we had to again improve it after revision. The lesson then is that developing analytical models can take a lot, lot of time, but it is always worth the while!

Little climatic niche overlap between Trachemys scripta and two of its congeners, T. venusta and T. cataspila. Figure credit: Sayra Espindola/Ella Vázquez-Domínguez

Describe the significance of this research for the general scientific community in one sentence.

The distribution, range limits and potential risk of the invasion of invasive species can be evaluated with genetic information and ecological niche modeling.

Describe the significance of this research for your scientific community in one sentence.

Evaluating interspecific interactions between native and non-native closely related species with genetic information and niche modeling approach was key to determine the distribution patterns, range limits and invasion risks of Trachemys scripta elegans.

Wetlands system in the valley of Cuatrocienegas, Coahuila, Mexico, where the endemic Trachemys taylori lives. Photo credit: Ella Vázquez-Domínguez

Espindola S, Vázquez-Domínguez E, Nakamura M, Osorio-Olvera L, Martínez-Meyer E, Myers EA, Overcast I, Reid BN, Burbrink FT. 2022. Complex genetic patterns and distribution limits mediated by native congeners of the worldwide invasive red-eared slider turtle. Molecular Ecology. https://doi.org/10.1111/mec.16356.

Interview with the authors: Associations between MHC class II variation and phenotypic traits in a free-living sheep population

Posted on by
Reply

In a recent paper in Molecular Ecology, Huang, Dicks and colleagues analysed variation in the major histocompatibility complex (MHC) and phenotypic traits in an unmanaged population of sheep living on an island off the coast of Scotland. This population of sheep has been studied closely for more around 70 years, providing a very rare level of insight and statistical power to evolutionary genetic studies. The MHC is among the most variable parts of mammalian genomes and has long been known to be encode proteins central to the adaptive immune system. Through their analyses, Huang, Dicks and colleagues found associations with levels of circulating antibodies and variation at MHC loci.

We sent some questions to the corresponding author of this work, Wei Huang, to get more detail on this new study.

Rams in St Kilda. Photo credit, Martin Adam Stoffel.

Can you describe the significance of this research for the general scientific community in one sentence?

This study demonstrated the direct link between immune genes and antibody levels in wild populations.

What led to your interest in this topic / what was the motivation for this study? 

The major histocompatibility complex (MHC) contains a number of genes linked with immune defence in vertebrates. Associations between MHC variation and phenotypic traits or pathogens have been identified in many species. Also, selection on MHC genes has also been demonstrated in some studies. However, many previous studies only examined associations between MHC variation and a limited number of phenotypic traits or pathogens. Few of them have examined both MHC-fitness associations and MHC-trait associations. The longitudinal study of Soay sheep in St Kilda is a great system to study the associations between MHC variation and phenotypic traits and how the associations are linked with selection on MHC genes. Using three representative phenotypic traits monitored in thousands of sheep over decades, we are able to provide a full picture of MHC-trait associations in wild populations.


Can you describe the significance of this research for your scientific community in one sentence?

This study suggests associations between MHC and phenotypic traits are more likely to be found for traits more closely associated with pathogen defence than integrative traits and highlights the association between MHC variation and antibodies in wild populations.

What difficulties did you run into along the way? 

It is extremely hard to monitor populations and collect longitudinal data over decades. Thanks to our great field assistants and volunteers, the Soay sheep data has provided a good foundation. In terms of the specific study, the first difficulty is to genotype MHC in a large number of sheep. We used two steps to genotype the MHC genes. We first used genotype-by-sequencing to genotype hundreds of sheep. Then, benefiting from the high-density sheep SNP chip, we were able to use 13 SNPs to genotype MHC in the other thousands of sheep successfully.

Additionally, it is hard to choose the appropriate model. Some of our traits are not normally distributed and are also not closed to other common error structures. We instead used Bayesian statistical methods to run the analysis.

What is the biggest or most surprising innovation highlighted in this study? 

We used three representative traits to examine the associations between MHC variation and phenotypic traits. The traits included a fitness-related integrative trait, body weight, a measure of gastrointestinal parasites, faecal egg count, and level of three antibodies. All of the three traits are related to fitness. We only found associations between MHC variation and antibodies. Such results reflect the important role of MHC in immune defence in wild populations. Our study is one of the first studies to examine associations between MHC variation and multiple phenotypic traits. 

How do you think your results generalize to other systems?
Our study is based on the longitudinal study of Soay sheep. The large sample size provides great statistical power. Therefore, our results are reliable and solid. Also, we investigated phenotypic traits that have different links with immune defence. Therefore, our results can reflect the general pattern of MHC-trait associations.  

You conclude from your study that MHC variation is more likely to be associated with immune traits. How would you validate your findings for species with less rich data?

First, it is possible to use experiments to test the associations. In terms of wild populations, future studies can investigate multiple populations or multiple traits in a single population if they are restricted by the study length.

Moving forward, what are the next steps in this area of research?

Our study demonstrates that it is important to study MHC-antibody associations. Future studies should focus on immune traits rather than only examine MHC-pathogen associations. Also, previous studies are often restrained by small sample size. It would be nice if future studies could increase their sample size to strength the statistical power.

Huang, W.*, Dicks, K. L.*, Ballingall, K. T., Johnston, S. E., Sparks, A. M., Watt, K., Pilkington, J. G., & Pemberton, J. M. (2022). Associations between MHC class II variation and phenotypic traits in a free-living sheep population. Molecular Ecology, 31, 902– 915. 

*These authors contributed equally to this work

Interview with the authors

Posted on by
Reply

A holobiont view of island biogeography: Unravelling patterns driving the nascent diversification of a Hawaiian spider and its microbial associates

In their recent paper in Molecular Ecology, Armstrong and Perez-Lamarque et al investigated the evolution of the holobiont. The holobiont is the assemblage of species associated with a particular host organism. In the case of this study, the holobiont refers to the stick spider (Ariamnes), its microbiome and its endosymbionts. Taking advantage of the successive colonization of islands in a volcanic archipelego, Armstrong and Perez-Lamarque et al contrasted the evolutionary history of the host species to the different components of the holobiont on different islands in Hawai’i.

We sent some questions to the authors of this work and here’s what Benoît Perez-Lamarque, Rosemary Gillespie and Henrik Krehenwinkel had to say.

Ariames waikula (on the island of Hawaii). Photo credit: George Roderick

What led to your interest in this topic / what was the motivation for this study? 

Gut microbiota play multiple roles in the functioning of animal organisms. In addition, host-associated microbiota composition can be relatively conserved over time and the concept of the “holobiont” has been proposed to describe the ecological unit formed by the host and its associated microbial communities. Yet, it remains unclear how the different components of the holobiont (the hosts and the microbial communities) evolve. This is what spurred our interest. Taking advantages of the chronologically arranged series of volcanic mountains of the Hawaiian archipelago, we were able to tackle this question and could investigate how the different components of the holobiont have changed as the host spiders colonized new locations.   

Can you describe the significance of this research for the general scientific community in one sentence?

The evolution of Hawaiian spider hosts and their associated microbes are differently impacted by the dynamic environment of the volcanic archipelago.
Can you describe the significance of this research for your scientific community in one sentence.

The host and its associated microbiota may not act as a single and homogeneous unit of selection over evolutionary timescales.

Ariames waikula (on the island of Hawaii). Photo credit: George Roderick

What difficulties did you run into along the way? 

All the different components of the holobiont are not as easy to study. For instance, for the host spiders, we used double digest RAD sequencing (ddRAD) to obtain genome-wide single nucleotide polymorphism data. With such data, we could precisely reconstruct the evolutionary histories of the different spider populations in the last couple of million years and tracked the finest changes in their genetic diversity. In contrast, characterizing the composition of the microbial components is much more challenging. We used metabarcoding of a short region of the 16S rRNA gene to identify the bacteria present. However, over such short evolutionary timescales, this DNA region is too conserved to accumulate many differences between isolated populations. Therefore, we had high-resolution data for the spider hosts but comparably low-resolution data for the bacterial communities. To ensure that the observed patterns were not artefactually driven by such differences of resolutions, we complemented our analyses with a range of simulations to assess the robustness of our findings.

What is the biggest or most surprising innovation highlighted in this study? 

We find that the different components of the holobiont (the host spiders, the intracellular endosymbionts, and gut microbial communities) respond in distinct ways to the dynamic environment of the Hawaiian archipelago. While the host spiders have experienced sequential colonizations from older to younger volcanoes, resulting in a strong (phylo)genetic structuring across the archipelago’s chronosequence, the gut microbiota was largely conserved in all populations irrespectively of the archipelago’s chronosequence. More intermediately, we found different endosymbiont genera colonizing the spiders on each island. This suggests that this holobiont does not necessarily evolve as a single unit over long timescales.

In the conclusion to your study, you point out how different components of the holobiont likely contribute differently to selection/colonization history in this system. If you had unlimited resources, what would you do to strengthen this conclusion? 

We indeed suspect that the different components of the holobiont probably did not act as a single and homogeneous unit of selection during the colonization of the Hawaiian archipelago. First, it would be ideal to perform an even broader sampling, targeting more Ariamnes populations and species from older islands, to better characterize the long-term changes of the different holobiont components. Using sequencing technics with better resolution (as detailed below) would also improve our characterization of the microbial component(s) of the holobiont. Second, to properly test for selection, we should perform transplant experiments of the bacterial communities between spider populations/species and measure whether or not it impacts holobiont fitness. We would expect to find a significant impact of the transplant for the endosymbionts, but no or low impact for the gut bacterial communities of these spiders.

The geological history of Hawai’i provides a powerful system to build understanding of the evolution of holobiont. Are you aware of other systems where similar studies could be performed? (I appreciate that this is related to the previous question!).

Many other archipelagos, with similar island chronosequences, like the Canary Islands or the Society Islands, are also ideal for testing hypotheses on the evolution of holobionts. Within the Hawaiian archipelago again, we could replicate our work on other holobiont systems. For instance, among arthropods, plant feeders might rely more importantly on their microbiota for their nutrition, and this might likely translate into different patterns of holobiont evolution.

Moving forward, what are the next steps in this area of research?

As previously said, one main limitation is the low resolution of the 16S rRNA metabarcoding. This prevented us to look at the evolutionary history of the individual bacterial lineages. Using a new model, we have recently tackled this issue of low resolution (https://doi.org/10.1128/msystems.01104-21) and we reported little evidence of microbial vertical transmission in these holobionts. Yet, the next step would be to move from classical metabarcoding to metabarcoding with longer sequencing reads (e.g. the whole 16S rRNA gene) or even metagenomics. It would provide more resolution for looking at bacterial evolution and would also bring more information on the functioning of these bacterial communities (e.g. are gut microbiota contributing to the digestion of these Hawaiian spiders in natural environments?).

Armstrong, E. E.*, Perez-Lamarque, B.*, Bi, K., Chen, C., Becking, L. E., Lim, J. Y., Linderoth, T., Krehenwinkel, H., & Gillespie, R. G. (2022). A holobiont view of island biogeography: Unravelling patterns driving the nascent diversification of a Hawaiian spider and its microbial associates. Molecular Ecology, 31, 1299– 1316. 

*Authors contributed equally