Choosy Snowcocks: How habitat preference can affect genetic population structure

Genetic study reveals striking differences in population structure for two snowcock species in the Himalayas.

The Tibetan Plateau houses two species of snowcock: the Tibetan Snowcock (Tetraogallus tibetanus) and the Himalayan Snowcock (T. himalayensis). These partridge-like birds occur in the mountain ranges around the plateau, but occupy slightly different habitats. The Tibetan Snowcock prefers high-altitude areas (above 3000 meters) with dwarf shrubs, while the Himalayan Snowcock can be found in low-elevation regions that are drier and warmer. These differences in distribution can affect their genetic make-up. A recent study in the journal Avian Research used several microsatellites and a mitochondrial marker to probe the population genetic patterns of these species.

Himalayan Snowcock

Himalayan Snowcock © Subharanjan Sen | Oriental Bird Club Image Database



Population Structure

The genetic analyses by Bei An and colleagues revealed some interesting differences between the two species. Populations of Himalayan Snowcock showed a “divergent and structured” pattern whereas there was no clear phylogeographic pattern in the Tibetan Snowcock. The researchers speculate that this difference is due to the distinct habitat preferences of these snowcocks.

The high-altitude lifestyle of the Tibetan Snowcock might render it more resilient to extreme cold. Hence, this species might have been affected less by the glacial cycles of the Pleistocene. Populations did not have to survive the cold periods in separate refugia and so do not exhibit any clear population structure. The Himalayan Snowcock, on the other hand, occurred at lower altitudes where populations were fragmented by the waxing and waning of the Pleistocene ice sheets. This explanation makes intuitive sense, but will need to be tested (for example with some ecological niche modelling).


Tibetan Snowcock © Donald Macauley | Wikimedia Commons



The mitochondrial marker indicated that Tibetan and Himalayan Snowcock are clearly distinct. The haplotype network, however, uncovered an intruiging insight. One haplotype (H9) was shared by both species. This finding suggests that there might be occasional hybridization. Indeed, both species co-occur in some regions, such as the Kunlun Mountains and the Qilian Mountains. More dense sampling is required to investigate whether these species actually interbreed.


The mitochondrial marker shows that both species are clearly distinct. However, one haplotype (H9 in the network) suggests that they might interbreed. From: An et al. (2020) Avian Research


An, B., Zhang, L., Wang, Y., & Song, S. (2020). Comparative phylogeography of two sister species of snowcock: impacts of species-specific altitude preference and life history. Avian Research, 11(1), 1.

Woodcreeper species continue to exchange DNA until 2.5 million years after divergence

Recent study estimates the time window for introgression in birds.

One of my favorite avian hybrids is the Swoose, a cross between a Greylag Goose (Anser anser) and a Mute Swan (Cygnus olor). This particular hybrid is probably sterile, because its parental species diverged more than 20 million years ago. It has been estimated that bird species can still produce viable offspring after, on average, 21 million years of independent evolution. This is considerably longer compared to mammals, where it takes about 4 million years for hybrid inviability to develop. The time window for hybridization has thus been established for birds and mammals, but what about introgression? How long can hybridization result in the exchange of genetic material between species? A recent study in the journal Evolution tackled this question using the neotropical genus Dendrocincla.




Paola Pulido‐Santacruz and her colleagues collected genomic data for 87 specimens, representing all six recognized species in this genus. First, they reconstructed the phylogenetic relationships between these taxa. The resulting evolutionary tree served as the backbone for a series of ABBA-BABA-tests to infer introgression. For readers unfamilar with this approach, I copied the explanation from a previous blog post (D-statistics for Dummies).

The rationale behind this test is quite straightforward: it considers ancestral (‘A’) and derived (‘B’) alleles across the genomes of four taxa. Under the scenario without introgression, two particular allelic patterns ‘ABBA’ and ‘BABA’ should occur equally frequent. An excess of either ABBA or BABA, resulting in a D-statistic that is significantly different from zero, is indicative of gene flow between two taxa. A positive D-statistic (i.e. an excess of ABBA) points to introgression between P2 and P3, whereas a negative D-statistic (i.e. an excess of BABA) points to introgression between P1 and P3.

The figure below illustrates this procedure for several Dendrocincla subspecies. In this example, there is an excess of ABBA-patterns, suggesting introgression between the subspecies neglecta and ridgwayi (belonging to the Plain-brown Woodcreeper, D. fuliginosa).


The ABBA-BABA-test was used to infer introgression between different Dendrocincla (sub)species. Here, an excess of ABBA-patterns points to introgression between the subspecies neglecta and ridgwayi (indicated with black arrows). From: Pulido-Santacruz et al. (2020) Evolution.


Time Window

These analyses uncovered five introgression events, which is probably an underestimate because the ABBA-BABA-test only captures introgression between nonsister taxa. Nonetheless, the researchers were able to date these introgression events, which ranged from “a few hunderd thousand to about 2.5 million years following divergence”. These results suggest that the timeframe for introgression is much narrower than the timeframe for hybridization. In Dendrocincla woodcreepers, the species boundaries seem to become impermeable after about 2.5 million years of separate evolution. Whether this time window is the same for other bird species remains to be tested.


The phylogenetic tree of the genus Dendrocincla with different introgression events indicated with colored arrows. From: Pulido-Santacruz et al. (2020) Evolution.


Genetic Snowballs

Another interesting observation in this study concerns the amount of genetic material than is exchanged between species. I explained the relevance of this pattern in a news article (a so-called digest) that accompanies the original study: “These proportions turned out to decline exponentially with the age of the hybridizing taxa. Interestingly, this pattern resembles the accumulation of genetic incompatibilities during the build-up of postzygotic isolation. These genetic incompatibilities do not increase linearly but instead seem to snowball because each new mutation is increasingly likely to be incompatible with a previous mutation.” However, this pattern is based on a few datapoints and will need to be confirmed with more data and other species.


The relationship between the proportion of DNA exchanged (fhom) and the age at hybridization follows a snowball pattern, reminiscent of the accumulation of genetic incompatibilities. From: Pulido-Santacruz et al. (2020) Evolution.



Ottenburghs, J. (2020). Digest: Avian genomes are permeable to introgression for a few million years. Evolution.

Pulido‐Santacruz, P., Aleixo, A., & Weir, J. T. (2020). Genomic data reveal a protracted window of introgression during the diversification of a Neotropical woodcreeper radiation. Evolution.

Drum roll, please! The evolution of rhythm in Woodpeckers

Large-scale analyses show how current selection on woodpecker drums might influence subsequent evolutionary trajectories.

A walk in the woods is often accompanied by the occasional drumming of a resident woodpecker. Experienced birdwatchers – or perhaps better: “bird-listeners” – can discriminate between the drums of different species. In fact, visualizing the different drum solo’s revealed that each species’ pattern conforms to a particular mathematical formula (see the figure below for some examples). The beauty of nature captured in a few symbols.

Similar to bird song, the drum rolls of woodpeckers serve a dual function. Males use these drums to attract females and to defend their territories against other males. These functions are the hallmarks of sexual selection. But which components of the woodpeckers’ drum – such as speed, length, cadence – are under sexual selection? A recent study in The American Naturalist compared the drums of about 200 woodpecker species to figure this out.

black woodpecker

The Black Woodpecker has a low, vibrating drum. © Alastair Rae | Wikimedia Commons


Character Displacement

This woodpecker-wide comparison revealed that species that live in the same area (i.e. they are sympatric) tend to produce drums with very different rhythms. This suggests that sympatric woodpecker species use rhythm to find a partner of their own species and avoid – possibly detrimental – hybridization. In technical terms, rhythm is undergoing sexual character displacement. Interestingly, the same is not true for the speed and length of a drum. These components are under directional selection and might thus be more important in territorial competition among males.


The variety of drums across the woodpecker phylogeny. The graphs on the right illustrate how different drums can be captured in a mathematical formula. From: Miles et al. (2019) The American Naturalist


Constraint and Potentiation

So, different rhythms allow the coexistence of certain woodpecker species. But these changes in rhythm have important implications for subsequent evolutionary trajectories. This is nicely illustrated by the Northern Flicker (Colaptes auratus) which contains two populations that might represent distinct species, namely the red-shafted (lathami) and yellow-shafted flicker (auratus). Individuals of the auratus population produce a linear cadence, while individuals of the lathami population have a constant rhythm. According to the estimates of this study, the lathami woodpeckers’ drum has the capacity to evolve more than ten times faster than the drum of auratus birds. In other words, the current drum rhythm “has introduced either a constraint on further signal evolution in auratus or a potentiation of further signal evolution in lathami.”


A Northern Flicker of the auratus population. © USGS | Wikimedia Commons



This phenomenon, where past selection influences the trajectory of future selection, is known as the recursive nature of evolution. This particular characteristic of the evolutionary process makes future evolutionary changes even harder to predict. Evolution is shaped by the interaction of deterministic and stochastic processes. Deterministic processes can be used to predict evolutionary outcomes, but the recursive feature of evolution can introduce stochastic elements in these deterministic changes, reshaping the evolutionary process in an unpredictable way.

You might be put off by this evolutionary quirk, but I think it only adds to the beauty of evolution. The authors capture this feeling nicely in their final sentence: “Altogether, out data illustrate how selection can be a source of contingency: it is a recursive process that modifies itself and, in doing so, leaves behind the exaptations and evolutionary ghosts that have fascinated biologists for decades.”



Miles, M. C., Schuppe, E. R., & Fuxjager, M. J. (2020). Selection for rhythm as a trigger for recursive evolution in the elaborate display system of woodpeckers. The American Naturalist, 195(5).

Did the New Zealand Rockwren ever occur on North Island?

A museum specimen suggests it did, but is this specimen reliable?

Nowadays, you can only find the New Zealand Rockwren (Xenicus gilviventris) on the South Island of New Zealand. This small songbird, in Māori known as pīwauwau (“little complaining bird”), is currently threatened with extinction, partly due to predation by introduced predators. To conserve this iconic bird, biologists are searching for predator-free areas to reintroduce it. One possibility is the other main island of New Zealand: North Island. However, conservationists prefer to reintroduce species into their former ranges, which raises the question: Did the New Zealand Rockwren ever occur on North Island?

Xenicus gilviventris

The New Zealand Rock Wren © Andrew | Wikimedia Commons


Museum Specimens

One way to figure out whether the New Zealand Rockwren ever hopped around North Island is to dive into museum collections. Alexander Verry and his colleagues found a specimen (labelled NHMUK 1939.12.9.75) that was collected by Henry H. Travers in the Rimutaka Ranges on North Island. Mystery solved!

No so quick (otherwise, this would be a very short blog post). Perhaps the specimen was mislabeled or it might represent a different species. To rule out these possibilities, the researchers extracted DNA from the museum specimen and compared it to modern samples with known origins.


The New Zealand Rockwren consists of a northern (red) and southern (blue) lineage. The specimen collected by Henry Travers (“Rimutaka Ranges”) clusters within the southern lineage. From Verry et al. (2019) Frontiers in Ecology and Evolution


Mistakes Happen

Previous genetic work revealed that the New Zealand Rockwren consists of two main groups: a northern and a southern lineage on South Island. The specimen collected by Travers clustered with samples from the southern lineage, a long way from the Rimutaka Ranges on North Island. It thus seems that the specimen was mislabeled.

Mistakes happen. However, mistakes appear to have happened quite regularly to Henry Travers. Three other specimens that he might have collected came from the Otago Province (according to the labels). The genetic data, however, suggest that these three samples originated from the northern part of South Island, some distance from the Otago Province. Moreover, Henry Travers was involved in other mislabeled specimens, such as skins from the South Island Snipe (Coenocorypha iredalei). It is reasonable to doubt the collection and labeling skills of Henry Travers.

The authors conclude that: “Our results suggest that New Zealand rock wrens have not been historically extirpated from New Zealand’s North Island, and that caution must be taken when utilizing museum specimens to inform conservation management decisions.”



Verry, A. J., Scarsbrook, L., Scofield, R. P., Tennyson, A. J., Weston, K. A., Robertson, B. C., & Rawlence, N. J. (2019). Who, Where, What, Wren? Using Ancient DNA to Examine the Veracity of Museum Specimen Data: A Case Study of the New Zealand Rock Wren (Xenicus gilviventris). Frontiers in Ecology and Evolution, 7, 496.

Let’s not hop to conclusions: Delineating species in the Rockhopper Penguin complex

There are three taxa in the Rockhopper Penguin complex. But are they all distinct species?

Just because you can differentiate between populations using genetic data doesn’t mean they should be considered separate species. The difficulty of delineating species was nicely discussed by Jeet Sukumaran and Lacey Knowles in a 2017 PNAS-paper. They showed that some methods for species delimitation – in their case the multispecies coalescent – mainly capture genetic population structure, but do not provide guidelines for drawing species boundaries. They concluded that “genomic-based results should only be considered a hypothesis that requires validation of delimited species with multiple data types, such as phenotypic and ecological information.”

I have expanded this conclusion in a book chapter, arguing that in order to delineate species, you need to understand the speciation process. This approach is nicely illustrated by a recent study in the Journal of Heredity that considers the taxonomy of Rockhopper Penguins (genus Eudyptes).


A Southern Rockhopper Penguin on Saunders Island, Falkland Islands © Ben Tubby | Flickr


Three Species?

The Rockhopper Penguin complex consists of two species: the Northern Rockhopper Penguin (E. moseleyi) and the Southern Rockhopper Penguin (E. chrysocome). The latter species contains two subspecies (chrysocome and filholi). However, some authors have suggested that these two subspecies should be elevated to species rank, because they can be separated based on several mitochondrial genes. But as I wrote in the beginning of this blog post: “Just because you can differentiate between populations using genetic data doesn’t mean they should be considered separate species.”

So, Herman Mays Jr. and his colleagues performed a phylogeographic study of the Rockhopper Penguins to gain more insights into their evolutionary history. This knowledge could then be used to better inform taxonomic decisions.


This haplotype network show that it is possible to discriminate between the three Rockhopper Pengiun taxa. But are they also different species? From: Mays et al. (2019) Journal of Heredity


Gene Flow

Analyses of the mitochondrial gene ND2 confirmed previous work: it is indeed possible to discriminate between the three taxa using mtDNA (see haplotype network above). But what about their evolutionary history? More detailed analyses indicated that Northern and Southern Rockhopper Penguins diverged about one million years ago and experienced a small (but significant) amount of gene flow. At around 0.5 million years ago, the two subspecies originated within the Southern Rockhopper Penguin. They showed high levels of gene flow.

These results can be interpreted in different ways. Currently, species concepts allow for some level of interspecific gene flow. But how much gene flow is too much? Perhaps some species delimitation software can help us out. A BPP (Bayesian Phylogenetics and Phylogeography) approach indicated that Northern and Southern Rockhopper Penguins should be considered separate species, while E. c. chrysocome and E. c. filholi do not qualify for an elevation to species rank.


Northern Rockhopper Penguin in Edinburgh Zoo, Scotland © Bardrock | Wikimedia Commons


Other Evidence

However, Jeet Sukumaran and Lacey Knowles already warned us not to take the results from species delimitation software at face value. What about other lines of evidence? The authors write the following:

We know of no evidence for separate species lineages in E. c. filholi and E. c. chrysocome outside of those suggested on the basis of mtDNA. There appear to be few if any fixed morphological, ecological, or behavioral differences between E. c. filholi and E. c. chrysocome outside of characters related to the coloration of the patch of skin at the base of the bill and any other phenotypic diversity that does exist among these taxa could likely be clinal.

More detailed studies are always welcome, but for now we can conclude that there are two species of Rockhopper Penguin.



Banks, J., Van Buren, A., Cherel, Y., & Whitfield, J. B. (2006). Genetic evidence for three species of rockhopper penguins, Eudyptes chrysocome. Polar Biology30(1), 61-67.

Mays Jr, H. L. et al. (2019). Phylogeography, Population Structure, and Species Delimitation in Rockhopper Penguins (Eudyptes chrysocome and Eudyptes moseleyi). Journal of Heredity110(7), 801-817.

Ottenburghs, J. (2019). Avian species concepts in the light of genomics. In Avian Genomics in Ecology and Evolution (pp. 211-235). Springer, Cham.

Sukumaran, J., & Knowles, L. L. (2017). Multispecies coalescent delimits structure, not species. Proceedings of the National Academy of Sciences114(7), 1607-1612.


This paper has been added to the Sphenisciformes page.

The peculiar genomes of Falcons

Falcons are the odd ball out when it comes to avian genomics.

Falcons are amazing birds. Take the Peregrine Falcon (Falco peregrinus), for instance. The large raptor can reach speeds over 320 km/h during a hunting dive, making it the fastest bird on this planet. From a hybrid point of view, falcons are also fascinating. Several hybrids have been reported in the wild, but most crosses are known from captivity. Falconers have combined certain species, aiming to create the perfect hunting bird. In the United Arab Emirates, for example, mixes of 7/8 Gyrfalcon (Falco rusticolus) and 1/8 Saker Falcon (Falco cherrug) are quite common. For an overview of hybridization in this bird group, you can check the Falconiformes page.

Clearly, there are plenty of good reasons to study falcons. But a recent review in the journal Ecology and Evolution adds another reason to this long list: genomics. It turns out that falcons are the odd ball out when it comes to bird genomes.


A Peregrine Falcon © Mosharaf Hossain | Wikimedia Commons


Some Taxonomy

Before we dive into the genomics of falcons, let’s see where they belong on the avian Tree of Life. Surprisingly, falcons are not closely related to other birds of prey. Instead, the closest relatives of falcons are parrots and songbirds, from which they diverged about 60 million years ago. More on this topic in this blog post.

Within the falcon family, you can find about 40 species. These can be divided into three main groups based partly on differences in hunting strategies. We already met the Peregrine Falcon which catches its prey in swooping dives. Other members of this first group of large and mid-sized falcons include Gyrfalcon and Sakerfalcon. The second group holds birds of the subgenus Hypothriorchis, better known as hobbies. These smaller falcons, such as the Sooty Falcon (Falco concolor), catch small birds and insects in flight. Finally, there are the kestrels, which often hover over grasslands, scanning the ground for unsuspecting prey.

Apart from these three main groups, there are several smaller groups, namely “1–2 species of merlins, two atypical African kestrels (Dissodectes), two Southern American falcons, the single New World Kestrel and two unplaced clades (“Red‐Footed Group” and “Southern Group”) each of which contains two species.” A wide range of falcons.


A Common Kestrel in flight © Вых Пыхманн | Wikimedia Commons



Now for some peculiar genomic falcon facts. So far, the genomes of five falcon species have been sequenced: Peregrine Falcon, Saker Falcon, Gyrfalcon, Common Kestrel (Falco tinnunculus) and Prairi Falcon (Falco mexicanus). Although these genomes have the typical length of an avian genome – about 1,200,000,000 nucleotides – they contain more and longer protein-coding genes compared to your average bird genome. These longer sequences cannot be explained by more repetitive DNA (such as transposable elements). Instead, some researchers think that this pattern is due to a lack of microdeletions, small DNA sequences (less than 30 nucleotides long) that have been removed from the genome. Indeed, the Peregrine Falcon had the lowest rate of microdeletions compared to seven other birds (see here).


Eurasian Hobby © Imran Shah | Wikimedia Commons



Mitochondrial genes can be found on – you guessed it – mitochondrial DNA, the small circular genomes in these power-producing organelles. However, several mitochondrial genes have been transfered to the nuclear genomes, giving rise to so-called NUMTs (nuclear mitochondrial DNA segment). These NUMTs take up a big part of falcon genomes. In fact, more than 90 percent of the Peregrine and Saker Falcon mitochondrial genome can be found in the nuclear genome, including one big insertion that represents 70 percent of the mtDNA (see here). These insertions might hold the ancestral DNA sequence of the falcon mtDNA and could be very useful in evolutionary studies.


African Pygmy Falcon © Sumeet Moghe |Wikimedia Commons



Finally, the falcon karyotype, the number of chromosomes you can count under a light microscope. In birds, chromosomes can be divided into a few huge macrochromosomes and several tiny microchromosomes. Most birds have 40 pairs of chromosomes (so 80 in total, mostly denoted as 2N = 80). Falcons, however, show karyotypes of only 2N = 40 to 2N = 52. These atypical chromosome counts are probably the outcome of fusions between microchromosomes into macrochromosomes.

These peculiar features – long genes, many NUMTs and atypic karyotypes – raise numerous questions. Hence, plenty of reasons to study falcons…


The karyotype of a Chicken. From: Li et al. (2015) BioMed Research International



Kapusta, A., Suh, A., & Feschotte, C. (2017). Dynamics of genome size evolution in birds and mammals. Proceedings of the National Academy of Sciences114(8), E1460-E1469.

Nacer, D. F., & do Amaral, F. R. (2017). Striking pseudogenization in avian phylogenetics: numts are large and common in falcons. Molecular Phylogenetics and Evolution115, 1-6.

Wilcox, J. J., Boissinot, S., & Idaghdour, Y. (2019). Falcon genomics in the context of conservation, speciation, and human culture. Ecology and Evolution, 9(24): 14523-14537.

Population Genomics of Mangrove Warblers: Looking for candidate genes in Costa Rica

Study on the mangrove warbler shows ecological speciation in action.

The American Yellow Warbler (Setophaga petechia) is extremely variable. This passerine, which breeds in the whole of North America and the northern part of South America, has been divided into 35 subspecies. Based on the breeding plumage of the males, ornithologists recognize three main groups: the golden warbler (petechia group, 17 subspecies), the mangrove warbler (erithachorides group, 12 subspecies), and the American yellow warbler (aestiva group, 6 subspecies). Hence, if you want to study this species, there is plenty of choice. A recent study in the journal Ecology and Evolution focused on a subspecies in the mangrove warbler group, namely xanthotera.


Ecological Speciation

This subspecies of mangrove warbler is distributed along an ecological gradient on the Pacific Coast of Costa Rica (not a bad place for fieldwork…). Loyal readers of this blog will probably know that such gradients are ideal locations to study ecological speciation. This process concerns the evolution of reproductive isolation between populations due to divergent natural selection along ecological gradients. As populations adapt to the changing conditions along these gradients, they diverge in certain characters which might result in the origin of new species.


A mangrove warbler in Costa Rica © Åsa Berndtsson | Wikimedia Commons


Some Predictions

At the start of an ecological speciation event, populations might already show some phenotypic differences. Despite these differences, there will be no clear genetic population structure, because the populations are connected by occasional gene flow (i.e. birds from neighboring populations can interbreed). However, this lack of genetic population structure is based on the whole genome. If you zoom in on particular genomic regions, you might find some genes that are slowly diverging. Which genes depends on the ecological gradient.

Tania Chavarria-Pizarro and her colleagues expected that differences in salinity and water availabilty might result in divergent selection on osmoregulation genes (i.e. regulation of salt concentrations in the body). Moreover, the size of prey also changes along the Pacific Coast, which could influence the evolution of bill morphology. Based on this background information, we can formulate some general predictions:

  1. There will be differentiation in some phenotypic traits, but no genetic differentiation.
  2. Genes involved in osmoregulation will be divergent
  3. Genes involved in beak morphology will be divergent



Genetic analyses confirmed the first prediction. The researchers did not detect any genetic population structure along the ecological gradient, suggesting that there are still high levels of gene flow between the populations. But is there differentiation in some phenotypic triats? This can be tested by comparing two statistics: Pst and Fst. Pst measures the differentiation in certain phenotypes, whereas Fst captures genetic differentiation. The first prediction is supported when Pst is significantly higher than Fst. And that was indeed the case for two traits: bill height and bill length!


Pst (grey line) is signficantly higher than Fst (red line) for bill height and bill lenght but not for wing length. From: Chavarria-Pizarro et al. (2019) Ecology and Evolution


Candidate Genes

The Pst-Fst patterns already indicated that there is divergent selection on beak morphology. But is this also reflected in the genome? To explore this issue, the researchers performed a genome scan, looking for divergent genes. This search uncovered 19 genes, of which several were involved in beak morphology and osmoregulation. This confirms predictions 2 and 3.

Let’s have a closer look at these candidate genes. In terms of osmoregulation, two aquaporins (AQP1 and APQ4) popped up. These genes code for proteins that regulate body water balance in various tissues. Very useful in a salty environment. Another gene involved in osmoregulation is NPNT, which is related to kidney development. The candidate genes for beak morphology were BMP1 and BMP5. These genes have also been linked to the evolution of beak morphology in Darwin’s Finches.


The genome scan revealed several genes under selection (red dots). Some genes are involved in osmoregulation (APQ1) and beak morphology (BMP5). From: Chavarria-Pizarro et al. (2019) Ecology and Evolution


Tip of the Iceberg?

This study nicely shows the environmental conditions are driving genetic and phenotypic differentiation in this mangrove warbler subspecies. This leaves still 34 other subspecies to be studied in greater detail, which will lead to new insights into the evolution of these songbirds. Indeed, the researchers conclude: “Further studies in other groups within S. petechia can lead to a better understanding the early stages of the formation of biological diversity in a group in which numerous populations could potentially constitute incipient or full biological species.” To be continued…



Chavarria‐Pizarro, T., Gomez, J. P., Ungvari‐Martin, J., Bay, R., Miyamoto, M. M., & Kimball, R. (2019). Strong phenotypic divergence in spite of low genetic structure in the endemic Mangrove Warbler subspecies (Setophaga petechia xanthotera) of Costa Rica. Ecology and Evolution, 9(24): 13902-13918.