Biology-MeFi: Chart of intra-species genetic diversity across phyla?

May 16, 2018 5:48 AM

I am looking for information on how genetic diversity within a species (not between species) differs between taxa. In other words: Is the "spread" in the genetic pool, if measured, for instance, in terms of single-nucleotide polymorphism (SNP), generally bigger / smaller for some groups of species than for others?

My ad-hoc idea is that species with a stronger phenotypic difference between individuals, like homo, should probably have a higher genetic diversity (higher rate of SNP) than those with lesser phenotypic variation, like, say, Orthoptera (grasshoppers etc.). But I couldn't find any statistics or research on this topic.

I am especially interested in data on far-related taxa. Ideally this would be a chart with comparative data for Sponges, Annelids, Arthropoda, Fish, Mammals (single representatives of these groups), or even Plants, Fungi, Archeae, Bacteria …

I do not look for information on the actual genome sequences, nor on biodiversity, but specifically on the genetic diversity within species from different phyla.

posted by megob to Science & Nature (18 answers total)

This chart from a talk by Richard Jovelin might be one starting point for you.

One interesting thing from that chart is that C. elegans has one of the lowest rates of genetic diversity found so far, while its cousin C. briggsae has one of the highest, with up to 20% of neutral sites being polymorphic.
posted by clawsoon at 6:25 AM on May 16, 2018 [2 favorites]

Right, I’ll look more later, but I think the chart can’t work the way you want it to, because within-group heterogeneity is going to swamp between-group variation. Even within genera, we have examples of extremely low and extremely high variation within species. It seems unlikely that within-species variation would be conserved for higher level taxa. At the extreme, many orders contain species that are mostly clonal.
posted by SaltySalticid at 6:32 AM on May 16, 2018 [2 favorites]

Molecular biologist in speculation mode, here. This is an interesting question, but is going to be complicated by the intrinsic taxonomic challenge of defining what a species is, and by what metric is used to assess genetic diversity. For instance, there’s a ton of microbe data that looks only at variations on the 16S ribosomal RNA, so I suppose in theory you could compare that to deep sequencing of the human small rRNA homolog, the 18S ribosomal subunit, but a) it’s still one gene and b) you still have that question of defining a species. Still, measuring biodiversity is an evolving field, but it’s not MY field, so it’s possible someone has or is in the process of addressing this. This paper looks like a good overview of some of the emerging methods and current challenges, and could give you some new terms for your search.

Also if I were generating a hypothesis about your question based on my cursory knowledge of genome architectures and evolutionary biology, it would be that species with more genome duplication events have greater within-species diversity. I believe that’s a general principle (albeit probably not taken to the individual variation level) in plants, many of which can have greatly increased chromosome number. Humans are diploid, with a chromosome number of 2 - two copies of each chromosome. Tomatoes have three. Wheat has a chromosome number of 6. Chromosome duplication is one way to dilute the cost of high genetic variation, since in principle, the more “backup” copies of a gene you have, the less penalty for deleterious impacts of genetic variation.
posted by deludingmyself at 6:33 AM on May 16, 2018 [2 favorites]

My ad-hoc idea is that species with a stronger phenotypic difference between individuals, like homo, should probably have a higher genetic diversity (higher rate of SNP) than those with lesser phenotypic variation, like, say, Orthoptera (grasshoppers etc.). But I couldn't find any statistics or research on this topic.

First: Are you sure that grasshoppers have a low rate of phenotypic variation?

Second: High rates of phenotypic variation can be achieved with no or little genetic variation; look at ant species with castes, for example, in which warriors and queens and workers have near-identical DNA but very different phenotypes.

Third: I'm trying to find another talk which discussed a related issue: Cell type diversity. A summary: Humans have a very low rate of genetic diversity, but we have a very high rate of post-translational modifications, especially in neurons. This doesn't give us high phenotypic variation, but it does give the opportunity to build a complicated brain. The chart in the talk which stood out showed a relationship between rate of post-translational modifications and brain complexity from worms to humans. So RNA diversity can lead to diversity in cell types, but not necessarily in phenotypes.

Fourth: The huge amount of phenotypic diversity in dogs - Yorkie to Great Dane - is mostly driven by half-a-dozen genes.

Fifth: Human genetic diversity is pretty low.

Sixth: Are you sure that human phenotypic diversity is high? There are morphs in many other species which blow away anything you see in humans.

Seventh: I got kind of carried away with my numberingth.
posted by clawsoon at 7:04 AM on May 16, 2018 [7 favorites]

Dog paper, since I missed the edit window: "In contrast to the results from association mapping of quantitative traits in humans and domesticated plants, we find that across dog breeds, a small number of quantitative trait loci (≤3) explain the majority of phenotypic variation for most of the traits we studied."

So my ad-hoc response to your ad-hoc idea is that it's probably wrong. :-) It's possible to build a lot of phenotypic diversity by having a single set of DNA that has highly flexible responses to environmental conditions (ant castes), or even has a randomness generator built-in. Life-cycle morphs - like caterpillar to butterfly, or blackback to silverback, or sequential hermaphroditism - are even more dramatic, and they occur with exactly the same DNA, in the same individual, throughout.

However, there may still be a weak association between genetic diversity and phenotypic diversity, so you may not be completely wrong.

If you're interested in the evolution of phenotypic variation in general, you might find Wallace Arthur's Evolution: A Developmental Approach interesting.
posted by clawsoon at 7:48 AM on May 16, 2018 [2 favorites]

So there's lots of problems with the question, the assumptions, and the framing. But that's not to say that it's not an interesting and valid question, it is! A few other comments that may help:

* genetic diversity within a population (let alone a species) is itself a difficult thing to study, and researchers use different methods that quantify different things. Sure, counting SNPs is one way, but by no means the only way, or even an especially meaningful way. A common tool used is RAPD (random amplification of polymorphic DNA), another is universally primed PCR (UP‐PCR). Both come under the heading of "genetic fingerprinting".

*Here is a paper that uses RAPD and UP-PCR to asses genetic variation in Timothy. The introduction has a good overview of how genetic diversity is/has been assessed, and points out problems with morphological and phenotypic approaches.

*Consider this revised question Is the "spread" in the genetic pool ... bigger / smaller for some groups of species than for others?. This can in principle be answered for specific groups via original research by you. Pick a selection criteria, find a bunch of papers that quantify diversity for some species or sets of species, and assemble the results. I'd think this would be a publishable meta analysis, but so far as I can tell it hasn't been done yet at the kind of taxonomic scale you're thinking of. So for instance just looking at a gscholar search for /RAPD genetic variation/ , I get 1.2k results, and see a dozen or so papers in the first few pages that would be interesting to compare. If you add a family of interest you might be able to meaningfully compare a small sample of species for each of a few genera, and come up with a very uncertain estimates of how a few genera compare.

*As a follow up on the clonal issue I raised above check out this paper , which compares triploid and diploid variants of an aquatic plant, each of which has different abilities regarding sexual and clonal reproduction.
posted by SaltySalticid at 8:22 AM on May 16, 2018 [3 favorites]

Response by poster: Bunch of very useful information here! And nothing is more useful than to have an ad-hoc hypothesis quickly rebutted. Just some quick (and totally incomplete) feedback / remarks:

– I found the paper by Richard Jovelin / Cutter / Day from which Jovelin took the chart. Indeed "hyperdiversity vs. near-clonal" seems to be the distinction I'm looking for. The Cutter Lab and its research is probably nearest to what I'm looking for.

– I am more interested in nuclear than in ribosomal/mito/plastid DNA, this can in fact be a (technical) problem. I will have a look into RAPD and PCR and try to figure out on what they are applied.

– human/grasshopper phenotypic diversity: this was even more ad hoc than ad hoc. But if we include ethological traits, I still think it could hold. I was looking for a general marker for "individualisation" or "individualizedness" (I'm a philosopher, you see ...)

– @clawsoon: rate of post-translational modifications – if you find the talk / article, it would be great to get a link. But if not, I suppose I'll find myself enough on the subject. Could be useful concerning question of biological base for "individualizedness"

– Life-cycle morphs ... are even more dramatic – yes. But here I am an adherer to the unorthodox hypothesis of serial larvae (Williamson). I know, I am quite alone with this. But again, I am a philosopher, speculation is our trade

Thank you all for your very diverse (!) and extremely helpful input!
posted by megob at 8:58 AM on May 16, 2018 [2 favorites]

Some ad-hoc ideas relating genotype diversity to phenotype diversity that you might be interested in exploring, in addition to what others have proposed above:

- Size of genome compared to complexity of organism. In pufferfish, there's hardly any "junk DNA", so I'd guess that most genotype changes will lead to a phenotype change. You can hardly have a mutation without bumping into something important. I'd guess the opposite for Paris japonica; with a genome that large, there are probably lots of places for mutations to happen that don't do anything. deludingmyself explained one common reason why.

- Size and age of population. A big, old population is going to have a lot of time to build up neutral mutations (e.g. in places like the third nucleotide of most codons). A recently bottlenecked population will not, even if evolutionary pressures in favour in phenotype diversity cause a small number of sites in the genome to quickly diversify. Examples of high variation in phenotypes based on a small number of genome sites: In the immune system, in domesticated dogs and pigeons, or in the snowflake yeast experiment in which you can dramatically transform the yeast phenotype within a couple of months.

- Type of evolutionary bet-hedging strategy used. A diversified bet-hedger will produce a wide range of phenotypes as offspring, while a conservative bet-hedger will produce a small range of flexible phenotypes. These strategies may be chosen independent of genomic diversity. I'd suggest that humans are examples of conservative bet-hedgers: We build a highly adaptable brain with relatively low phenotypic variation on top of very low genotypic variation, and it's able to deal with a huge number of environments and cultures. We don't need either high genomic or high phenotypic diversity, because we've gone all-in on conservative bet-hedging.

- Sex or cloning. Sex allows high genomic diversity to develop at spots which must vary quickly (e.g. the beaks of finches) while keeping it low for genes which you don't want to vary (e.g. RNA transcriptase). (Bacteria trade plasmids to get a similar advantage.) If you're purely cloning, your genetic variation is going to be the result of random mutations and is just as likely to hit something critical, something junk, or a useful variation.

Since you're a philosopher, here's a philosophical question which has an impact on the answer: In phenotypic variation, are you including phenotypes which can't survive? If a genetic mutation leads to a phenotype which dies in utero, does that count a phenotypic variation? There are whole set of developmental considerations which may or may not apply to your question depending on your answer.

Alright... I should go to my real job now. [sigh]
posted by clawsoon at 9:15 AM on May 16, 2018 [2 favorites]

megob: But if we include ethological traits

This leads to an even more interesting philosophical question. :-) Does learning count as a change to the phenotype?

When the ducklings imprinted on ethologist Konrad Lorenz, were they phenotypically different from duckling who did the normal thing and imprinted on their mother?

Is a child who learns German phenotypically different from a child who learns English?

If your answer is yes, you might be interested in my earlier question about all of the different ways that biological systems can learn. Some of them involve DNA; some of them involve DNA being manipulated in weird and wonderful ways; some of them involve neurons; some of them involve completely different systems.
posted by clawsoon at 10:23 AM on May 16, 2018

Thank you all for your very diverse (!)

If you keep digging further and further into this, one thing that you'll find is that if there's more than one way to do something, nature will do it all of those ways. And then it will come up with a few more ways to do it that you never would've thought of. That's part of why the responses are so diverse.

Phenotypic diversity is one of those things which nature achieves by every possible means. Genomic variation is only one of them.
posted by clawsoon at 11:01 AM on May 16, 2018

Response by poster: @clawsoon,

– are you including phenotypes which can't survive – no. My question is aimed at accumulated (or accumulable) variations, be it phenotypic or genotypic

– learning: beyond the scope of my current reflection here. But your question is interesting, although "accumulating knowledge" seems to stem from an information-theoretical approach, while your answers (and those from other mefites) are mostly of adaptationist order. But sure, if your subject of knowledge is "biological system", both do conflate. I think I have something I can add to the list, will check this later.

– diversity of ways of diversification: Sure. Nature is exuberance.
posted by megob at 11:35 AM on May 16, 2018

are you including phenotypes which can't survive – no. My question is aimed at accumulated (or accumulable) variations, be it phenotypic or genotypic

In that case: Developmental complexity might be the most important determinant of phenotypic variation. If DNA is building a tree, it's no problem to have 10 limbs or 100 limbs. If DNA is building a mouse, anything other than 4 limbs is a big problem.
posted by clawsoon at 11:55 AM on May 16, 2018

Here's a philosophical issue regarding diversity that you may want to ponder. First, note that diversity of genotypes within population isn't really mathematically or conceptually any different from species diversity in a place. Both are concepts of biodiversity, and biodiversity is what you are interested in, albeit one aspect of it that is not as famous among the general populace as "whoa there are a million species in the Amazon". So with that out of the way, I will pretend we're talking about plants on farms.

Consider two farms.
farm A has 95 acres of oats, and one acre each of wheat, rye, corn, soy and barley.
farm B has 20 acres each of oats, wheat, rye, corn and soy.

Which is more diverse? You may say A, because it has 6 crops, while B has only 5. But if you get dropped into farm A at random, you're pretty much only going to see oats, while in B, you'll see lots more diversity on average. So B is more diverse in that sense. What about other farms:
farm C has 20 acres each of oats, wheat, rye, corn and soy, but two of the soy types are a little different
farm D is the union of farms A and B

Should D have diversity equal to A, or B, or should it be greater, or perhaps even less than either? How does C compare to A or B?

What group is more diverse depends on how you define and quantify it, and how you define it (should) depend on why you want to quantify diversity, i.e. what's your goal here. Because there is no one definition that is always best.

Diversity is far more than a simple count of different types. We speak of species richness, (aka alpha diversity), beta diversity, gamma diversity, and also things like evenness and Shannon's H. There are many definitions of diversity, and they all have their uses. All of these concepts have straightforward analogues for genetic diversity.

So before you try to get data, it is probably worthwhile to figure out which aspect of diversity you are interested in, and why that notion will be useful to your bigger picture, as opposed to any of the alternatives.
posted by SaltySalticid at 1:41 PM on May 16, 2018 [3 favorites]

I've been working on a side project on comparative genetic diversity, and have done a bunch of reading recently. It seems like genetic diversity is related to certain life-history traits, but not strongly connected to population size, genome size, or ploidal level. In fact, given the range of population sizes, thought to be an important factor in genetic diveristy, the range of observed diversity levels is shockingly narrow. This problem is known as Lewontin's Paradox.

Here are some papers you might find useful:

This one's PloS, so open-access:
Leffler et al. 2012. Revisiting an Old Riddle: What Determines Genetic Diversity Levels within Species?
Figure 1 might be what you're looking for.

A couple of relevant reviews, in Nature publiciations, so paywalled, sorry.
Ellegren & Galtier. 2016. Determinants of genetic diversity
Figure 1 in particular.

Romiguier et al. 2014. Comparative population genomics in animals uncovers the determinants of genetic diversity
Figure 1, in particular.

I am willing to provide copies of these for the purposes of scientific discourse! And happy to discuss further.
posted by vortex genie 2 at 1:53 PM on May 16, 2018 [4 favorites]

Great links vortex genie, thanks! Last I looked into this in any detail was around '13, but the Leffler paper does seem like it's pretty close to what I said may not exist yet. I very much appreciate how those works broadly support this:
genetic diversity... not strongly connected to population size, genome size, or ploidal level.
At a skim, it seems that we can add phyla and many ranks of high level taxonomical organization as 'not playing a large role in determining within-species genetic diversity'.
posted by SaltySalticid at 5:41 PM on May 16, 2018

Response by poster: vortex genie – these are exactly the charts and the discussions I was looking for, even Porifera is there! The articles confirm my suspicion that the question is worth to be asked (again), and they do this by disproving my initial naive idea. That's how it should be.

Interesting to see these two correlations, one with fecundity rate, the other with habitat (aquatic / terrestrial). As to habitat, one more unorthodox idea comes to my mind, which could even have some realistic core: As the concentration of viruses is much higher in water than in air, viral transfections could play a role in the higher genetic diversity of marine organisms.

(In parentheses: My own research on scholar did not bring up these articles ... Human reading/research experience shows out to be still much superior – this is good news! And thanks for your offer to send the papers – I have institutional access here and got them easily. Will read on now, not ready yet for further discussion, but will get in touch happily whenever I am!)
posted by megob at 3:08 AM on May 17, 2018

I'm a champion lurker, but just happen to be writing a review about This Very Topic. I read your question, megob, and thought, "I read a stack of papers about that exact thing." So I had to chime in.

It's really cool what's happening as tons of genomic data are accumulating. So many fascinating evolutionary, functional, and comparative questions are being investigated, some of which date back to the modern synthesis.

I've read TONS of literature on this recently, so please feel free to get in touch for discussion or questions or more reading recommendations :)
posted by vortex genie 2 at 12:54 PM on May 17, 2018 [1 favorite]

vortex genie 2: I'm a champion lurker, but just happen to be writing a review about This Very Topic.

Oh. Man. When you finish, I need to know.
posted by clawsoon at 1:25 PM on May 17, 2018