Rather than give a long introduction to hybrid speciation, I refer you to a recent post I did on diploid hybrid speciation in the Galápagos finches; just have a look at the introduction, which talks about the commonness of hybrid speciation in plants (via polyploidy) and its rarity in animals. The Galápagos finches may be a case of diploid speciation following hybridization, but it’s not clear, for they don’t get a chance to mate with one of the two ancestral species (i.e., we don’t know if it’s reproductively isolated from one parent), and it’s also likely, I think, that this small-population hybrid species won’t last very long, but will be “mated to death” by its one sympatric parent.
But there’s a new report of a hybrid species in manakins from the Amazon, and this one looks pretty good. Published in PNAS (reference below, free access, and pdf here), its first author is Anfredo Barerra-Guzmán, and the last author is Jason Weir, a postdoc in Trevor Price’s lab who was next door to me for a few years.
The paper is long and complex, with some genetic details that aren’t necessary for general readers, but I’ll summarize the results as succinctly as I can.
In the Brazilian Amazon, at the headwaters of the Cururu-ri River in Pará state, exist three species of manakins: Lepidothrix vilasboasi, L. nattereri, and L. iris, which has two subspecies. L. vilasbosi, which I’ll call LV, lives in an area intermediate between L. nattereri (“LN”) and L. iris (“LI”); all are separated by rivers, which for many Amazonian birds constitute geographic barriers because birds simply don’t like to fly over water. Here are their geographic distributions from the paper:

You can see that LV, the putative “hybrid species” lies in between its two parents LN and LI (two subspecies). When LV was first found in 1957, was thought to be simply a “hybrid zone”: an area of hybridization between the two parental species.
The three species differ most markedly in the color of their head crowns. As the paper notes:
The crown patch in L. iris is iridescent and varies from brilliant white (its usual look, which is very similar to L. nattereri) to blue or purple, depending on the angle of light. Males of the two subspecies of L. iris distributed on either side of the Xingu River are almost identical in plumage, with L. iris iris possessing a thin green strip between the upper mandible and the crown patch and with the crown patch extending all the way to the mandible in L. iris eucephala. Females (not shown) appear like males but lack the contrasting crown and rump patches and do not differ appreciably among species.
In contrast, LV has a yellow crown patch, very distinct from that of the two parental species (see below).

Here’s LV (the “golden-crowned manakin”) with its yellow crown:

And LN (the “snow-capped manakin”) with its white crown:

Well, what is the evidence that LV was formed after hybridization between LN and LI? That rests on genetics. The genome of LV is a mixture of genes from the two parental species, and attempts to reconstruct the ancestry of these birds, using both standard phylogenetics and “coalescent” simulations, show that it’s far more likely that LV came from mixed genes from LN and LI than that it branched off from one or the other species (the conventional “branching” scenario for speciation). Between 15% and 38% of the LV genome comes from LN and 62%-85% of the genome from LI, depending on what method is used to do the calculation.
One problem is that LV could still represent an isolated hybrid population, or a hybrid zone, between the two species. The authors say this is unlikely because pure F1 hybrids of the two species (that is, first-generation offspring), would have 50% of the genes from each species rather than this skewed distribution. But that would be the case only if the F1 hybrids are sterile so that they couldn’t form a “hybrid swarm” by further reproduction. It’s still possible that we have here a hybrid zone or swarm in which members are not reproductively isolated from the two parental species, so that LV is not really a “species”. The crucial evidence is whether LV is reproductively isolated from LN and LI (note: it’s already geographically isolated, but that doesn’t count as a reproductive barrier in the “biological species concept”, for in that case any geographically isolated population would be a species).
So is LV reproductively isolated from the two parents? We don’t know for sure, for they don’t coexist in the same area. Barrera-Guzmán et al. suspect that LV would be isolated—by its differently colored crown. Crown colors are used by birds as ways to identify mates from their of own species, and to discriminate against the wrong ones, so the authors suppose that LV males would not be recognized as appropriate mates by females from LN and LI because of LV’s novel yellow crown. Likewise, it’s possible that LV females wouldn’t recognize the iridescent crowns of its parental species, so there would be reproductive isolation on all fronts.
The interesting thing about the LV crown is its novel yellow color. The authors found that the iridescences of the LN and LI crowns is caused by different structural properties of their feathers, and that the LV crowns are intermediate between these two sets of properties. The interesting thing is that if you make a pure F1 hybrid between LN and LI, you get a dull crown with the same intermediate structural properties of the putative hybrid LV, but it’s not yellow!
The authors then speculate—reasonably in my view—that what happened in this scenario was that the two parental species hybridized, producing a population that lives in the forest (where LV does) but had a dull crown. Sexual and/or natural selection then caused carotenoid-retaining structures to evolve in the hybrid population so that males would be visible to females in the forest. (Female preferences for yellow would, by sexual selection, go along with this.) Thus, hybridization itself wasn’t sufficient for speciation: the reproductive isolation would have to have involved some post-hybridization selection.
The upshot: I think the authors have a good case for having uncovered a rare diploid hybrid species of bird. It’s not an absolutely compelling case, as it’s still possible we have a hybrid swarm that is not a new species and is not reproductively isolated from the parents. If they could demonstrate strong mate discrimination among species based on crown color (or anything else), the data would be more convincing. But I still think they’ve uncovered a possible (even probable) case, which, with some follow-up work, could represent one of just a handful of diploid hybrid species in animals.
Some people may say, “well, this is a new evolutionary paradigm, for speciation is supposed to occur by branching of lineages, not by anastomosis (fusion) of separate branches into one.” And yes, if this is the way species normally formed, it would mandate a pretty big revision of Darwinian theory. But the evidence is that this is not the way most diploid species form, for if it did, phylogenetic analysis would not resolve any bits of the tree of life: we’d just get a spaghetti-like mess. In Drosophila, the group I know best, this is clearly not the case: branching rather than fusion is the rule. Although there is more leakage of genes between animal species than we suspected two decades ago, that is “introgresssion”, not speciation. We still see little evidence that the formation of new species in animals occurs by the fusion of already-existing branches (species).
______
Barrera-Guzmán, A. O., A. Aleixo, M. D. Shawkey, and J. T. Weir. 2018. Hybrid speciation leads to novel male secondary sexual ornamentation of an Amazonian bird. Proc. Nat. Acad. Sci. USA 115:E218-225. Published ahead of print December 26, 2017, doi:10.1073/pnas.1717319115