Two species produce viable hybrids even though they diverged 150 million years ago

Here’s an amazing paper—though the striking part isn’t sufficiently emphasized—reporting artificial hybridization between two long-diverged species, and the resulting appearance of viable hybrids despite what must be substantial genetic divergence between the components of a hybrid genome.

Here are the two species involved, the Russian sturgeon (Acipenser gueldenstaedtii), a source of meat and caviar, and a member of the fish family Acipenseridae.

From Wikipedia

The other species is the American paddlefish (Polyodon spathula), a member of the fish family Polyodontidae, and the only living species in that genus (the Chinese paddlefish, Psephurus gladius, which is in a different genus, has been pronounced extinct).   Note the different morphology: the “paddle” on the fish below, which it uses for filter feeding (the Russian sturgeon is a predator on crustaceans, molluscs, and smaller fish), and the lack of bumpy protuberances on the paddlefish’s back. The fin shapes and placement also differ. These species are strikingly divergent in morphology, behavior, and, of course, in time.


How long since these two species had a common ancestor? If you go to the great site Timetree, which gives divergence times of many species calculated from the literature (you can enter any pair and will usually get an answer), you find a median divergence time of 136 million years and an estimated time of about 158 million years.  That is a very long time! (The paper below, using older data, says the divergence time is longer: about 184 million years.)

For comparison, the divergence between our species and the Virginia opossum occurred about 160 million years ago, so the sturgeon/paddlefish hybrids are equivalent, with respect to divergence time, to getting a viable hybrid between a human and a possum!


The hybridization is reported in the new paper below from the journal Genes (click on screenshot to go to paper and get the pdf; reference at bottom).

The results can be stated briefly. During an experiment designed to get Russian sturgeon offspring without a father’s genes (“gynogenesis”), the researchers did a control cross between female Russian sturgeon and male American paddlefish (sturgeon eggs were put in a solution of diluted paddlefish sperm). There were five such crosses. They showed a hatching rate of 78-85%, and a survival rate of hybrids at 6 months from 49-68%—comparable to survival in artificial fertilization experiments in pure species.

One complication: Russian sturgeon are “polyploid”, that is, their ancestors underwent a complete duplication of their genomes, so they have four copies of each chromosome (i.e., they are “tetraploid”). Most modern fish are polyploid in this way. Paddlefish, however, diverged so long ago that they are “diploid,” that is, like most modern vertebrate species, they have only two copies of each chromosome.

Because of this difference, two kinds of hybrids were produced. “SH hybrids” were “triploids”, with three sets of each chromosome: two from the sturgeon egg and one from the paddlefish egg (total chromosomes around 165). They also obtained “pentaploid” hybrids, called LH, apparently resulting from the fusion of a sturgeon egg that had a full set of maternal chromosomes plus one set of paddlefish chromosomes (total chromosomes around 300).

You can see the two types of hybrids below. (a) is the Russian sturgeon, and (the caption is wonky), the rest of the pictures show hybrids, without a picture of the paddlefish itself). (c) is the LH hybrid with a full sturgeon genome (four copies of each chromosome) and half a paddlefish genome. That’s why it looks mostly sturgeon-y, with a bit of a paddle in front.

(d) is the “LH” triploid hybrid, which has half a sturgeon genome less than (b). That’s why it looks more paddlefish-y.

I’m assuming that these pictures are labeled correctly in the paper; but it could be that (b), unlabeled in the graph, is an SH hybrid, (c) the LH hybrid (it has a longer “paddle”), and (d) the pure paddlefish. If there’s a labeling error, it should be fixed.

Regardless, the surprising thing is that viable and apparently vigorous hybrids were produced. Were they fertile? No data on that are reported, but, given the difference in ploidy level between the parents, I doubt it: the different numbers of chromosomes from each species would play hob with meiosis—the chromosome-pairing process that is required for the formation of sperm and eggs.

But still, despite about 150 million years apart, the genomes can still work together to produce a viable and functional individual. That is simply amazing; like I said, it’s sort of like getting a viable offspring between a human and a possum. (What a monstrosity that would be: a furry anthropoid with huge teeth and a prehensile tail!)

But why could such long-diverged species still have their genes function fairly harmoniously in a single individual? Well, we don’t know, but there are several possibilities.

1.) While the two species could be long diverged, the function of the diverged genes hasn’t changed much, so they don’t screw up development. In other words, adaptive evolution in these species has been slow since they shared their common ancestor. (This isn’t true for non-coding bits of the genome, for that’s the way they estimated the great divergence time of the species.) I’m not that keen to embrace this possibility since the fish are so different in appearance and behavior, and in ways that seem adaptive.

2.) The species have an “open” developmental system that will tolerate errors or disharmonies more readily, compensating for them by producing a functional individual. This used to be the explanation for why animals like frogs could produce viable hybrids between long-diverged species, while mammals couldn’t. But saying the developmental system is “open” is just another way of saying that “long-diverged genes still function together”, and is thus a bit tautological. To really test  whether the individual species are more tolerant of screwups, one could, I suppose, look at the effects of mutations in these species that usually wreck development in fish. If sturgeon and/or paddlefish are less disrupted by such mutations than are other fish, one might have something to go on.

3.) The polyploidy of the sturgeon genome somehow buffers development against screwup. That is, having two or four copies of the sturgeon genome in hybrids can somehow override the detrimental effects of genome incompatibility.  I’m not sure exactly how this would work, but perhaps it would be through a “dosage” effect: if you have a full dose (at least two copies) of one species’ genes, you can override any effects of other-species’ genes that act to “poison” development.

Well, we don’t know the answer. What is remarkable is the resilience of the genome, and the development it induces, in the face of ancient divergence. Remember, this is about 22 times the temporal divergence between humans and chimpanzees, and we cannot produce hybrids with chimps (it’s been tried; see Why Evolution is True).


Káldy, J., A. Mozsár, G. Fazekas, M. Farkas, D. L. Fazekas, G. L. Fazekas, K. Goda, Z. Gyöngy, B. Kovács, K. Semmens, M. Bercsényi, M. Molnár, and E. Patakiné Várkonyi. 2020. Hybridization of Russian Sturgeon (Acipenser gueldenstaedtii, Brandt and Ratzeberg, 1833) and American Paddlefish (Polyodon spathula, Walbaum 1792) and Evaluation of Their Progeny. Genes 11(7), 753;


  1. jezgrove
    Posted July 14, 2020 at 10:20 am | Permalink

    Astonishing! The Timetree site looks interesting, too.

    • Simon Hayward
      Posted July 14, 2020 at 12:00 pm | Permalink

      Timetree also has an app – certainly for iPhones I assume android too

      • Mark Joseph
        Posted July 14, 2020 at 11:59 pm | Permalink

        I’ve got the Timetree app; it’s useful for those bizarre middle-of-the-night thoughts such as “when did humans and bananas diverge?” (just under a billion and a half years ago; probably somewhat less in the case of Ray Comfort).

    • Torbjörn Larsson
      Posted July 14, 2020 at 6:05 pm | Permalink

      Timetree is great, and they have been updating – but of course YMMV.

  2. GBJames
    Posted July 14, 2020 at 10:30 am | Permalink


  3. ThyroidPlanet
    Posted July 14, 2020 at 10:31 am | Permalink


    • ThyroidPlanet
      Posted July 14, 2020 at 1:28 pm | Permalink


  4. Desnes Diev
    Posted July 14, 2020 at 10:44 am | Permalink

    I hope the researchers will test if the hybrids are fertile. Reproduction in fish is sometimes surprising (from a mammalian perspective).

    • Simon Hayward
      Posted July 14, 2020 at 12:03 pm | Permalink

      Yes. I never knew there were these ploidy (which autocorrects to “policy”)issues in fish. It feels more like plant than animal breeding now.

  5. Posted July 14, 2020 at 11:28 am | Permalink

    I understand that the rate of evolution is studied. But does that take into account one might say “degrees of resulting incompatibility”?

  6. Posted July 14, 2020 at 11:41 am | Permalink

    Would it work with paddlefish eggs and sturgeon sperm, I wonder.

    • Posted July 14, 2020 at 11:43 am | Permalink

      Good question. Surprising that the experimenters didn’t try that.

  7. darrelle
    Posted July 14, 2020 at 12:10 pm | Permalink

    Amazing indeed! The human – possum example really puts it in perspective.

    Has any general trend of polyploid species being more successful at producing viable hybrids than diploid species been noted in any research?

    • Susan H.
      Posted July 15, 2020 at 12:07 pm | Permalink

      In vertebrates, it’s sort of the other way around–all the successful polyploid species I know of are hybrids, but most viable hybrids are diploid. All-female triploid hybrid species of guppies and whiptail lizards are well known, they reproduce by parthenogenesis.

      • darrelle
        Posted July 15, 2020 at 12:40 pm | Permalink

        Thank you Susan.

  8. Ken Kukec
    Posted July 14, 2020 at 12:24 pm | Permalink

    … equivalent, with respect to divergence time, to getting a viable hybrid between a human and a possum!

    Boy, it would take a long time between intraspecific assignations, and a much-more-attractive-than-average possum, even to give that one a try.

    • rickflick
      Posted July 14, 2020 at 1:10 pm | Permalink

      It might be hard, even in a Petri dish. Just the thought of what you were doing…1960s horror films flash before your eyes.

      • Ken Kukec
        Posted July 14, 2020 at 6:41 pm | Permalink

        Hell, even most farm boys draw the line at carnal knowledge of members of the family Didelphidae.

        • jezgrove
          Posted July 14, 2020 at 6:47 pm | Permalink

          Yup, but only “most” – “Grab ’em by the possums” is doubtless the refrain of the others…

          • rickflick
            Posted July 14, 2020 at 7:04 pm | Permalink

            What a nightmare.

  9. BobTerrace
    Posted July 14, 2020 at 12:27 pm | Permalink


  10. Rasko
    Posted July 14, 2020 at 12:28 pm | Permalink

    Thanks for the TimeTree link, PCC(E). Hadn’t seen this one. The last time I did a phylogenetic analysis, I had to estimate divergence times using a pencil and a ruler on a copy of a published tree, then make up a NEXUS file from my estimates. Exceedingly tedious. But now I can calibrate my NEXUS file for the next set of analyses.
    Am very grateful.

  11. Malcolm
    Posted July 14, 2020 at 12:35 pm | Permalink

    I was reading this paper yesterday, fascinating stuff!

  12. Steve Pollard
    Posted July 14, 2020 at 12:48 pm | Permalink

    The internet works in mysterious ways! Only 24 hours after PCC(E) was lamenting the dearth of interesting science papers to write about, along comes this little beauty. Utterly fascinating; I hope the researchers (or others) develop a serious, testable hypothesis about what is going on.

  13. Jacques Hausser
    Posted July 14, 2020 at 1:35 pm | Permalink

    I think that the labels are indeed mixed in the figure’s legend. The “d” individual looks like a plain paddlefish, and it seems quite logical to arrange the display like that: maternal species, hybrids, paternal species.
    If this hypothesis is ture, it is interesting to note that the hybrids look quite sturgeon-like, even “c” with its long nose. They are sturgeon species with a long nose, e.g. A. stellatus. It could be that the paddlefish genome is practically silenced in the hybrids.
    The tetraploidy of A. gueldenstaedtii is probably itself the memory of a past hybridization.

    • Jacques Hausser
      Posted July 14, 2020 at 1:37 pm | Permalink

      true, not ture… 😦

    • Posted July 14, 2020 at 4:05 pm | Permalink

      I was wondering about this too. That could go far in explaining why the hybrids look more sturgeon-y, and why this crazy experiment even worked at all.
      Great stuff, though!

    • Torbjörn Larsson
      Posted July 14, 2020 at 6:08 pm | Permalink

      I was going to comment something like that on image and ploidy – silencing didn’t occur to me – so: agreed.

    • Posted July 14, 2020 at 6:44 pm | Permalink

      I agree with you, Jacques; I think the figure is badly mislabeled.

    • Posted July 14, 2020 at 6:55 pm | Permalink

      I thought the same thing about the “sturgeon-ness” of the hybrids. Pure strength in numbers on the chromosome battlefield? Or is it more sophisticated than that?

    • Borys Wrobel
      Posted July 15, 2020 at 5:24 am | Permalink

      …yes it would be interesting to see to what extent the two genomes are transcriptionally active in the hybrids. Perhaps the lack of such an analysis is the reason this paper has not been published in a journal with a higher renown.

  14. Hempenstein
    Posted July 14, 2020 at 4:01 pm | Permalink

    Same number of chromososomes in each species? Genomic sequences of each known?

  15. Alex K.
    Posted July 14, 2020 at 5:35 pm | Permalink

    A shame the Chinese paddlefish was declared extinct last year. The potential to study its biology in comparison to its other acipenseriform relatives is now lost forever except for a few museum specimens.

    The American paddlefish is the last member of its clade now that originated from the Early Cretaceous.

  16. Posted July 14, 2020 at 5:50 pm | Permalink

    Great post. Quick question: does it matter for the ability of the hybrids to reproduce whether the sturgeon are allotetraploid or autotetraploid? (I.e. whether the four copies of each chromosome pair as two pairs or one fourtuple?) And whether the three homologous chromosomes in the hybrid all pair together? My guess is that you get into trouble as a result of meiosis in the hybrid either way.

    • Posted July 14, 2020 at 6:46 pm | Permalink

      That’s my guess, Joe.

    • Jacques Hausser
      Posted July 14, 2020 at 7:26 pm | Permalink

      I too doubt that the three homologous chromosomes can “pair” together. The only way such hybrids could produce viable gametes would be through a mecanism similar to the hybridogenesis in the frogs of the genus Pelophylax, where the edible frog (P. kl. esculentus) is a hybrid between the pool frog (P. ridibundus) and the marsh frog (P. lessonae). In esculentus, recombination doesn’t occur and it produces only lessonae gametes. Thus if the esculentus mate with a ridibundus partner, they produce a further generation of hybrids, that is esculentus again; if they cross with lessonae, they produce pure lessonae. The “real” situation is a tad more complex because some esculentus are triploid and can produce both types of gametes.

    • Posted July 14, 2020 at 8:05 pm | Permalink

      In the paper, the sturgeon are described as “functional tetraploids”, which to me suggests that during meiosis they form fourtuples, and thus that they are autotetraploids, not amphidiploid allotetraploids. Spontaneous autopolyploidy is apparently common in sturgeon; see

  17. jezgrove
    Posted July 14, 2020 at 6:51 pm | Permalink

    BBC Radio 4’s The Curious Cases of Rutherford and Fry had a good recent programme about hybrids last week:

  18. kelskye
    Posted July 14, 2020 at 8:12 pm | Permalink

    Wow, that’s an incredibly unexpected result. Can’t wait for a deeper analysis to explain why such divergent species are viable.

  19. gravelinspector-Aidan
    Posted July 17, 2020 at 1:42 pm | Permalink

    Verily 2020 will go down in the annals as “The Year of Who Ordered That?”

    (What a monstrosity that would be: a furry anthropoid with huge teeth and a prehensile tail!)

    Coming soon to the late night bars at a “furry” convention …

  20. Posted July 18, 2020 at 2:47 am | Permalink

    Seems I remember watermelons having an interesting genetic arrangement like this to make them seedless.

  21. Posted July 28, 2020 at 2:46 pm | Permalink

    Just read a somewhat sad update on the fate of hybrids:

    “…The Times reports that approximately 100 sturddlefish remain alive today. The team behind the accidental hybridization says that they have no plans for creating more sturddlefish in the future—a short-lived life for a species that was never meant to be.”

%d bloggers like this: