Applying the Biological Species Concept to Bacteria

Introgression is not limited to Eukaryotes.

Over the years, I have written several blog posts about species concepts (see for example here and here). I argued that most biologists currently follow the General Lineage Concept or the Evolutionary Species Concept, which both regard species as independently evolving lineages. Laypeople are probably most familiar with the Biological Species Concept (or BSC), defining species as “a group of organisms that can successfully interbreed and produce fertile offspring.” However, this concept can be difficult to apply in certain situations, such when populations are geographically isolated and will never meet. Another common criticism is that the BSC cannot be applied to Bacteria because they do not reproduce sexually. You can imagine my surprise when I came across a recent paper in the journal Genome Biology where researchers applied the BSC to Bacteria:

Some bacteria can engage in gene flow via homologous recombination and this observation has led a growing number of researchers to suggest that bacterial species and speciation might be best defined using the same evolutionary theory developed for sexual organisms; the biological species concept (BSC).

Introgression and Recombination

Awa Diop and her colleagues studied more than 30,000 bacterial genomes. First, they classified these genomes into species by using a cut-off value of 94% genetic similarity in a set of core genes. This arbitrary threshold is commonly applied to delineate bacterial “species”. In this study, however, it mainly allowed the researchers to create a set of “species” for further analyses. To determine whether there has been introgression between these different bacterial “species”, the researchers calculated the ratio between homoplasmic (h) and non-homoplasic (m) alleles. A homoplasmic allele is a genetic variant that is not the result of inheritance from parent to offspring (i.e. vertical inheritance). Instead, such an allele can be the outcome of introgression between bacterial species (i.e. horizontal transfer) or convergent evolution (i.e. bacteria that independently acquire the same mutation). Clonal species – that reproduce asexually – are expected to have few homoplasmic alleles and thus a low h/m ratio. Introgression will result in an increased h/m ratio due to the accumulation of homoplasmic alleles.

In addition, introgression will be accompanied by recombination, the exchange of homologous sections of chromosomes. This process leads to the breakdown of linkage between certain alleles – also known as linkage disequilibrium (LD) – across chromosomes. Clonal species don’t engage in recombination and will thus show no reduction in linkage disequilibrium.

The researchers simulated bacterial genomes without gene flow and compared these patterns – in terms of h/m ratio and LD – with the actual data. These analyses revealed that most bacterial “species” showed signs of introgression and only 2.6% were truly clonal. Some kind of sexual reproduction among Bacteria seems to be more common than we expected.

Genomic analyses pointed to high levels of gene flow (or introgression) between bacterial species. From: Diop et al. (2022).


Although the level of introgression among bacterial “species” varied extensively (see figure above), it correlated nicely with sequence similarity. The more similar two species are on a genetic level, the higher the level of introgression uncovered in this study. This pattern can be explained by the observation that homologous recombination requires nearly identical stretches of DNA (also known as MEPS, Minimal Efficient Processing Segments). As genomes diverge, the density of these MEPS decreases and recombination becomes less likely. The relationship reported in this study shows a rapid reduction in introgression between 2% and 10% of sequence divergence. This result explains why an arbitrary threshold to define species of about 95% has been so useful in the past. However, introgression occurred between species that were 90% to 98% divergent. The exact threshold for bacterial species boundaries will thus depend on the study system. There is no silver bullet.

The relationship between sequence identify and level of introgression shows a sharp turn at ca. 90% sequence divergence. From: Diop et al. (2022).

From Bacteria to Birds

You might be wondering why I am covering a paper about bacterial species on a blog dedicated to birds. There are two main reasons: (1) I have a broad interest and don’t want to limit myself to literature on avian hybridization, and (2) you can learn a lot from other study systems. In this case, I noticed an interesting parallel between the arbitrary species threshold in birds (ca. 2% divergence in mitochondrial genes) and in Bacteria (ca. 95% divergence in core genes). These thresholds can be useful as a starting point, but are not always reliable (see for example this blog post). Moreover, this study confirmed a growing consensus among biologists studying speciation: introgression is more common than we previously thought. It doesn’t matter whether we are talking about Bacteria or birds.


Diop, A., Torrance, E. L., Stott, C. M., & Bobay, L. M. (2022). Gene flow and introgression are pervasive forces shaping the evolution of bacterial species. Genome Biology23(1), 1-19.

Featured image: Neisseria gonorrhoeae © Dr. Norman Jacobs | Wikimedia Commons

What caused the decline of the Green Peafowl?

Did past climatic changes or human actions impact this species?

In the current climate of rapid biodiversity loss, it is easy to blame human activities. Land use changes, overexploitation or other anthropogenic factors can certainly contribute to population declines, but in some cases past climatic changes have left their mark. Humans just provided the final push over the edge, hurling the population towards extinction. Disentangling the impact of past climate change and recent human-induced impact is a challenging exercise. Luckily, the methodological toolbox keeps expanding. And these tools are even more powerful when they are combined in an efficient way.

In a recent study, a team of Chinese scientists applied several approaches to understand the downfall of the Green Peafowl (Pavo muticus), an endangered bird species from Southeast Asia. They published their findings in the Proceedings of the Royal Society B.

Demographic Patterns

First, the researchers used genomic data to reconstruct the fluctuations in the population size of the Green Peafowl (you can check out this blog post for more details on the specific method they used, namely a PSMC analysis). This demographic analysis revealed an early population decline between 800,000 and 210,000 years ago, followed by a recovery during the Last Interglacial Period (about 70,000 years ago). After this period, the population started declining again. Unfortunately, it is not possible to determine whether this decline continued into the present day. The results of a PSMC analysis become unreliable in more recent times.

That is why the researchers turned to another approach and sequenced the genomes of five museum samples (from 1956 to 1976). Comparing the genetic make-up of these older specimens with present-day birds pointed to a significant reduction in genetic diversity. It thus seems that the decline in population size from the glacial periods can be extended to the present day.

The demographic PSMC analysis indicated a steady population decline until about 10,000 years ago (left). A comparison between museum specimens and modern samples revealed a significant decrease in genetic diversity, suggesting that the population decline has continued until the present day. From: Dong et al. (2021).

Niche Models

Using genomic data, we have now established that the Green Peafowl has been declining since the last ice age. But we still don’t know whether humans were involved. To answer this question, the researchers took another look at their toolbox. They reconstructed the amount of suitable habitat during the Holocene (less than 10,000 years ago) with Ecological Niche Modelling (ENM). This approach “predicted stationary general range during these periods and imply little impact of climate change.” If we can rule out these climatic changes, it had to be anthropogenic factors. Right?

Not so fast. This type of reasoning would be a black-and-white fallacy (i.e. pretending that there are only two options). To confidently blame human actions, we need more direct evidence. Here, the researchers provided two lines of evidence. First, they reported a negative correlation between human disturbance statistics, such as intensified land use for buildings and agriculture, and the population size of the Green Peafowl. In addition, they referred to written records in Chinese history that described the use of meat and feathers from this bird species.

Several indices of human impact increased over time, such as population size (blue line), area with buildings (orange), cropland (yellow) and grazing land (purple). These variables negative correlate with Green Peafowl population sizes over time. From: Dong et al. (2021).

Shades of Grey

Taken together, it seems reasonable that human factors have played a central role in the decline of the Green Peafowl. Nonetheless, I would argue that the reductions in population size during the Pleistocene might have rendered this species more vulnerable for population decline in more recent times. As mentioned in the previous paragraph, we should not fall victim to a black-and-white fallacy. The world is often comprised of different shades of grey. Safeguarding the future of the Green Peafowl will add some much-needed color.


Dong, F., Kuo, H. C., Chen, G. L., Wu, F., Shan, P. F., Wang, J., … & Yang, X. J. (2021). Population genomic, climatic and anthropogenic evidence suggest the role of human forces in endangerment of green peafowl (Pavo muticus). Proceedings of the Royal Society B288(1948), 20210073.

Featured image: Green Peafowl (Pavo muticus) © Scaup | Wikimedia Commons

Slow down, please: The evolution of beak morphology in Tanagers

Which evolutionary model best explains the evolution of this bird group?

The early bird gets the worm. This saying not only applies to our everyday life, it can also be relevant for evolution. When a species colonizes a new area, its members might be confronted with numerous vacant ecological niches. Some individuals might adapt to feed on worms, while others prefer grains or fruits. This situation of ecological opportunity sets the stage for rapid diversification and the origin of new species. In other words, an adaptive radiation. From a theoretical point of view, you would expect an initial burst of species diversification followed by slowdown of evolutionary changes as the niches are being filled.

This scenario has been described for island populations, but does it also apply to species that spread across continental landmasses? A recent study in the Biological Journal of the Linnean Society tested this model with such a group of species: the tanagers. About 12 million years ago these birds colonized South America and diversified into more than 300 species with a wide range of beak morphologies. The ideal study system to explore the early burst scenario on a large spatial scale.

Three Models

Nicholas Vinciguerra and Kevin Burns collected data on the beak morphology for 333 out of 377 species of tanagers. Next, they summarized all morphological variation in a few metrics and compared three different evolutionary models to explain the observed variation:

  • Brownian Motion (random changes over time)
  • Ornstein–Uhlenbeck (evolution towards an optimal value)
  • Early Burst (the model described above)

The analyses revealed that the Early Burst model was the best-fitting model. The researchers noted “a rapid burst of bill shape evolution early in the evolutionary history of tanagers followed by a subsequent slowdown toward the present.” This finding supports the scenario that tanagers quickly filled the available morphospace in beak morphology when the ecological opportunities were present. Over time, the available niches filled up and the rate of evolutionary change dropped.

The phylogenetic analyses showed that an increase in new species (blue line) is accompanied by an early burst in beak morphology followed by a slowdown in evolutionary diversification (black lines). From: Vinciguerra & Burns (2021).


More detailed analyses revealed that the Early Burst model also applied to specific subfamilies, namely the core tanagers (Thraupinae), the highland tanagers (Diglossinae), the warbler tanagers (Poospizinae), the saltators (Saltatorinae) and the honeycreepers and allies (Dacninae). Adaptive radiations nested within a larger adaptive radiation. Similar patterns have been found in Vangas after they colonized Madagascar, but the situation in the tanagers appears more extreme.

However, the accumulation of species and morphological disparity within vangas occurred 23 Mya within an insular system, whereas in tanagers this evolution has occurred on a continental scale in nearly half the amount of time.

Interestingly, the Darwin’s Finches – the textbook example of an adaptive radiation – did not follow the Early Burst model. The lack of this iconic subfamily in the list above can probably be explained by their recent evolutionary origin. The Darwin’s Finches are still in the early stages of an adaptive radiation. If we could wait a few thousands to millions of years, we might see a slowdown in evolutionary rate in these birds.

An overview of the different subfamilies within the tanagers. Five of these groups also showed an Early Burst pattern of diversification in beak morphology.


Vinciguerra, N. T., & Burns, K. J. (2021). Species diversification and ecomorphological evolution in the radiation of tanagers (Passeriformes: Thraupidae). Biological Journal of the Linnean Society133(3), 920-930.

Featured image: Purple honey creeper (Cyanerpes caeruleus) © Charles J. Sharp | Wikimedia Commons

How different are Mallards and Chinese Spot-billed Ducks on a genetic level?

A recent study detected minor differences on the sex-chromosomes.

Morphologically, Mallards (Anas platyrhynchos) and Chinese Spot-billed Ducks (A. zonorhyncha) are easy to tell apart. First of all, the sexes of the Mallard are drastically different whereas male and female Chinese Spot-billed Ducks look alike. In addition, the Chinese Spot-billed Duck can be recognized by its pale head which is marked by a whitish eyebrow and two black stripes. And it sports a yellow spot on the bill from which it derives its name. Interestingly, these morphological differences do not extend to the genetic level. Analyses of mitochondrial DNA and several nuclear markers could not discriminate between these species.

The observation of clear morphological disparity without genetic divergence is not uncommon in birds. I have covered several cases on this blog, such as redpolls and warblers. A mismatch between morphology and genetics can often be explained by a few differentiated genomic regions that underlie the phenotypic differences. Hence, Irina Kulikova and her colleagues took another look at the genetic make-up of the Mallard and the Chinese Spot-billed Duck. Did they find any genetic differences?

Genetic Outliers

The researchers scanned the genomes of 23 Spot-billed Ducks, 29 Mallards and 3 hybrids. In the end, they obtained more than 3000 genetic loci: 3130 on the autosomes and 194 on the Z-chromosome (i.e. one of the sex-chromosomes in birds). Most of the genetic variants at these loci were shared between the two species, confirming previous work that they are genetically similar. However, genetic differentiation was about 4.5 times higher on the Z-chromosome compared to the autosomes. A more detailed look at this sex-chromosome revealed three loci that were significantly different between Mallard and Chinese Spot-billed Duck. Moreover, these loci popped up when the researchers tested for signatures of divergent selection. There are thus genetic differences between these duck species. We just had to look really hard to find them.

The Z-chromosome is highly differentiated between Mallards and Chinse Spot-billed Ducks. It contains three clear outlier (depicted as triangles) that might underlie the morphological differences. From: Kulikova et al. (2022).

Future Work

Finding the genetic differences between these duck species is only the first step. Now, the researchers want to find out whether these genetic outliers directly contribute to the morphological differences that we observe. We know that genes regulating plumage coloration and bill color often reside on the sex-chromosomes (see this review by Darren Irwin). Mallards and Spot-billed Ducks might be another example. But this hypothesis remains to be tested with more fine-scale genomic analyses. Nonetheless, the researchers are confident that they are on the right track:

Whether these regions are involved in phenotypic differences between the species and sexual dimorphism is the prospect of future work. We believe that whole-genome sequencing along with plumage analyses will shed light on phenotypic evolution and help to identify speciation mechanisms in Mallard and Chinese Spot-billed Duck.


Kulikova, I. V., Shedko, S. V., Zhuravlev, Y. N., Lavretsky, P., & Peters, J. L. (2022). Z‐chromosome outliers as diagnostic markers to discriminate Mallard and Chinese Spot‐billed Duck (Anatidae). Zoologica Scripta.

Featured image: Chinese Spot-billed Duck (Anas zonorhyncha) © Alpsdake | Wikimedia Commons

Big geological events and a small bird: the evolutionary story of the Rifleman

Can its current distribution be explained by past or recent events?

“Species that appear widespread may in fact be represented by a series of isolated sub-populations with varying demographic and diversity characteristics, and their genetic and phenotypic diversity may therefore be at higher risk of extinction than they first appear.” This statement caught my attention while I was reading a recent Ecology and Evolution paper. In this study, Sarah Withers and her colleagues investigated the population structure of the Rifleman (Acanthisitta chloris) on the North Island of New Zealand. This small bird has limited dispersal capabilities due to its reduced tail and wing morphology. We could thus expect to find many isolated sub-populations of this species across the island. And indeed, the researchers noted that the North Island Rifleman (subspecies granti) is currently limited to a highly fragmented distribution. This observation raises another question: how did this patchy distribution originate? Is it a consequence of past geological changes or recent human-induced impacts?

Three Lineages

To answer this question, the researchers examined the genetic population structure of the Rifleman with two mitochondrial markers (COI and the control region) and a set of twelve microsatellites. The genetic analyses revealed three mitochondrial lineages: an insular population (Little Barrier Island), a southeastern population (Tararua Ranges and Mohi Bush), and a mainland population (the remainder of birds on the island). The microsatellites largely corroborated this division.

But when did these three lineages arise? Based on a mitochondrial molecular clock, it turned out that the insular population split off from the mainland between 1.25 and 2.5 million years ago. This timing corresponds to the rising sea levels during the Pleistocene epoch, resulting in a significant barrier between the mainland and the islands. Surprisingly, the divergence between the southeastern populations and the rest of the mainland was much older, dated to ca. 4.9 million years ago. These results indicate that the current distribution of the Rifleman has been largely shaped by past geological events. Let’s explore these events in more detail below.

A principal coordinates analysis of the microsatellites points to three distinct genetic lineages. From: Withers et al. (2021).

Drawing Lines

The divide between the southeastern and the mainland populations coincides with two biogeographical barriers: the Taupo Line and the Cocknaye’s Line. The first line (Taupo Line) can be explained by the formation of the Manawatu Strait which divided the North Island into a northern and a southern section. The second line (Cocknaye’s Line) follows a mountain range that arose during the Pleistocene, pushing populations into eastern and western regions. Based on their findings, the researchers suggest that “the distribution of genetic diversity within and among North Island Rifleman populations contains the genetic signal of both the Taupo Line and Cockayne’s Line.” Probably, the Rifleman were first divided into northern and southern populations along the Taupo Line. Later on, the southeastern populations were isolated from the rest of the mainland along the Cockayne’s Line. The figure below provides a graphical overview of these events.

The geographical events explaining the current distribution of the Rifleman on North Island. Populations were first divided across the Taupo Line (figure B), followed by further subdivision along the Cockayne’s Line (figure C). From: Withers et al. (2021).

A Special Case

And there you have it: the biogeographical history of the Rifleman. Interestingly, similar patterns have been documented in non-avian taxa, such as insects and plants. But not in birds. The Rifleman seems to be an exception. Indeed, the researchers write that “Rifleman therefore appear to be a remarkable case among birds in that they show the genetic signal of past dispersal barriers on a scale usually seen only in invertebrates and plants.” A bird species with a non-avian history.


Withers, S. J., Parsons, S., Hauber, M. E., Kendrick, A., & Lavery, S. D. (2021). Genetic divergence between isolated populations of the North Island New Zealand Rifleman (Acanthisitta chloris granti) implicates ancient biogeographic impacts rather than recent habitat fragmentation. Ecology and Evolution11(11), 5998-6014.

Featured image: Rifleman (Acanthisitta chloris) © Christopher Stephens | Wikimedia Commons

What determines range shifts up and down tropical mountains?

Exploring the impact of different ecological traits.

Some bird species might be on an “escalator to extinction.” As the climate changes and the suitable habitat shifts upslope, the birds have no choice but to move along. At some point, however, the mountain stops and the species might go extinct. This scenario makes intuitive sense, but is obviously too simple. Several other factors play a role in elevational range shifts and not all species will move up the mountain. Indeed, several studies reported that between a third and a fifth of species actually shift downslope.

In a recent Frontiers in Ecology and Evolution paper, Montague Neate-Clegg and his colleagues compiled a dataset of elevational range shifts for 421 bird species. They found that the species tend to move upslope with an average speed of 1.6 meter per year. However, this speed is only an average which does not capture the variation in underlying range shifts. These shifts are influenced by several ecological traits. Let’s head up this mountain of data and explore.

Species Traits

The dataset contained information from eight study sites across the tropics, from Peru to New Guinea. Detailed statistical analyses revealed that “elevational shift rates are associated with species’ traits, particularly body size, dispersal ability and territoriality.” The finding that dispersal ability plays a role in range shifts is not that surprising. Birds that disperse farther are more likely to explore new locations outside of their current distribution. Similarly, the influence of territoriality is also logical. Birds that do not hold territories are more free to move around and colonize new areas. Finally, the relationship between body size and elevational shifts can be explained within the context of life history theory. Small-bodies species shifted their ranges faster. Smaller species tend to have faster life histories, allowing them to rapidly respond to changing environments. Hence, we have three factors – dispersal ability, territoriality, and body size – that make ecological sense.

The analyses showed that body mass and territoriality have an effect on elevational shifts in tropical birds. These graphs illustrate the relationships with mean elevation (A and B), lower limit (C) and upper limit (D) of the distribution. From: Neate-Clegg et al. (2021).


As mentioned above, not all bird species move upslope. And indeed, this study also found that a third of the shifts were downslope. What could explain this reversed movement? The authors offer two main ideas: (1) competitor release and (2) tracking of ecological requirements. The first explanation entails the scenario that a competitor goes extinct downhill, allowing another species to extend its distribution into the freely available niche and extend its range downslope. The second explanation is quite intuitive: species move where there is food. If your favorite food source happens to be down the mountain, that is where you will go there.

In the end, this study provides several ecological features to understand the movement of tropical bird species up and down mountains. However, they did not explicitly consider large-scale factors, such as the topography of the landscape and the configuration of the vegetation. A nice opportunity for further research into elevational range shifts. Much remains to be discovered, so I guess it is not all downhill from here.

The percentage of species that shifted downslope. The different shared of grey correspond to mean elevation, upper limit and lower limit. From: Neate-Clegg et al. (2021).


Neate-Clegg, M. H., Jones, S. E., Tobias, J. A., Newmark, W. D., & Şekercioǧlu, Ç. H. (2021). Ecological correlates of elevational range shifts in tropical birds. Frontiers in Ecology and Evolution9, 621749.

Featured image: Andean Motmot (Momotus aequatorialis) © Alejandro Bayer Tamayo | Wikimedia Commons

Adaptive introgression between two high-altitude duck species

Genetic analyses suggest exchange of hemoglobin genes.

Last week, Svante Pääbo was awarded the Nobel Prize in Physiology or Medicine “for his discoveries concerning the genomes of extinct hominins and human evolution.” Together with many colleagues, he discovered how Neanderthals and Denisovans have contributed to the evolutionary story of humans. Hybridization appears to have been quite common, leading to genetic exchange among archaic humans and these extinct species. Building on this work, Emilia Huerta-Sánchez and her colleagues documented introgression from Denisovans into Tibetans, allowing the latter to adapt to life at high altitudes. Specifically, the gene EPAS1 – which plays an important role in dealing with low oxygen conditions – was transferred between these ancient hominins. A beautiful example of adaptive introgression.

Similar examples have been reported in animal species that live at high altitudes, such dogs and cattle (see this blog post). However, these cases do not feature humans and thus received less attention from the media. Luckily, there are some science websites – such as this blog – that put the spotlight on some hidden gems in the vast scientific literature on hybridization. Recently, Allie Graham and her colleagues investigated whether adaptive introgression also occurred in two South American duck species: the Speckled Teal (Anas flavirostris) and the Yellow-billed Pintail (A. georgica). Their findings were published in the journal Heredity.


Both duck species are widespread across South America and can be found at high altitudes in the Andes. These mountain populations are thus interesting study systems to understand adaptation to high altitude. Given the examples of adaptive introgression in other high-altitude animals and the high incidence of hybridization in ducks, it seems reasonable to look for evidence of adaptive introgression in these two species. The researchers focused on 31 genes that are involved in the production of hemoglobin and the physiological reaction to low oxygen conditions (i.e. the HIF pathway).

A commonly used test for introgression is the D-statistic, also known as the ABBA-BABA test. 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. Applying this approach to the two duck species revealed significant D-statistics for the β-globin genes, but not for any of the other genes in the study (i.e. α-globin genes and the HIF pathway).

Calculating the D-statistic indicated an excess of shared alleles between the high-altitude populations of the Yellow-billed Pintail and the Speckled Teal. This pattern can be due to introgression. From: Graham et al. (2021).

Convergent Evolution?

However, a significant D-statistic does not necessarily mean introgression. The same pattern can be the outcome of other evolutionary processes, such as convergent evolution. Perhaps the high-altitude populations of Yellow-billed Pintail and the Speckled Teal independently acquired the same mutations in the β-globin genes? To rule out this explanation, the researchers took a closer look at the genomic region containing these genes. Interestingly, the β-globin gene cluster showed very low genetic differentiation between the high-altitude populations. This reduced differentiation was not limited to the genes, but extended across the whole genomic region. In addition, a phylogenetic network of the β-globin genes revealed a clearly separate cluster of haplotypes that contained introgressed alleles that were identified in a previous study. Together, these patterns indicate that the β-globin gene variants are not the result of parallel mutations, but are identical by descent. Introgression is thus the most likely explanation.

Two lines of evidence suggest that the β-globin gene cluster was introgression. The top figure shows that the genomic region has low genetic differentiation between the high-altitude duck populations. And the lower figure reveals a separate cluster (highlighted in red) with the introgressed genetic variants. Adapted from Graham et al. (2021).

Faster Evolution

All in all, this study provides convincing evidence for introgression of β-globin genes between the two duck species. Moreover, demographic modelling revealed that the genes flowed from Speckled Teal into the Yellow-billed Pintail. This finding allowed the researchers to sketch a possible scenario:

Thus, the yellow-billed pintail received these variants through hybridization and may not have waited for de novo mutations to adapt to the high-altitude environment, but rather acquired beneficial alleles from the standing variation, albeit via another species with a similar genetic background, leading to faster evolution.

Although adaptive introgression seems likely, it remains to be explicitly tested if the exchange of these genes was indeed adaptive. One possible analysis could be performed with the so-called VolcanoFinder. This approach scans the genome for certain genetic patterns that look like volcanos when plotting pairwise genetic differences. In such an analysis, the β-globin gene cluster should be a clearly visible peak, similar to the Andean mountains where these duck species reside.


Graham, A. M., Peters, J. L., Wilson, R. E., Muñoz-Fuentes, V., Green, A. J., Dorfsman, D. A., Valqui, T. H., Winker, K. & McCracken, K. G. (2021). Adaptive introgression of the beta-globin cluster in two Andean waterfowl. Heredity127(1), 107-123.

Featured image: Yellow-billed Pintail (Anas georgica) © Brian Ralphs | Wikimedia Commons

Is the Altai Snowcock a hybrid species?

Re-analysis of genetic data questions the hybrid origin of this species.

Hybrid bird species are rare. In my review of hybrid speciation in birds, I listed only seven putative cases with variable levels of supporting evidence. Since the publication of that paper, a few other bird species have been proposed to be of hybrid origin, namely the Salvin’s Prion (Pachyptila salvini) and the Steller’s Eider (Polysticta stelleri). It is, however, difficult to discriminate between hybrid speciation and other evolutionary scenarios, such as incomplete lineage sorting and repeated bouts of introgressive hybridization. Several lines of evidence are needed to confidently conclude that a bird species has a hybrid origin (see for example, the Italian Sparrow [Passer italiae]). That is why I am always very skeptical when someone announces the discovery of a hybrid bird species. As Carl Sagan nicely put it: “extraordinary claims require extraordinary evidence.”

Hybrid Snowcocks

Recently, Li Ding and colleagues proposed that the Altai Snowcock (Tetraogallus altaicus) evolved through ancient hybridization between Tibetan Snowcock (T. tibetanus) and Himalayan Snowcock (T. himalayensis). This conclusion was largely based on a phylogenetic analysis of the mitochondrial D-loop which clustered three hybrid individuals next to the Tibetan Snowcock. Molecular dating analyses suggested that these hybrids shared a common ancestor with the Tibetan Snowcock about 1.8 million years ago. The researchers summarized their conclusions at the end of the paper:

The hybridization between T. tibetanus and T. himalayensis reproduced fertile hybrids during the Quaternary glacial period and repeatedly backcrosses with T. himalayensis during the interglacial period as a result of inheriting many characteristics from T. himalayensis, and glacial dispersal and isolation finally promoted the speciation of T. altaicus.

That is quite a bold statement from a single evolutionary tree. As mentioned above, I am always skeptical about claims of hybrid bird species. And I am not the only one: Martin Päckert was not convinced by the presented evidence and decided to re-analyze the genetic data. His results appeared in the journal Ecology and Evolution.

Phylogenetic Analyses

The hybrid species hypothesis of the Altai Snowcock relies heavily on the phylogenetic position of two mitochondrial haplotypes: H35 and H36. In the original analyses, these haplotypes formed a separate cluster next to the Tibetan Snowcock. However, Päckert could not recover this result. He ran several phylogenetic analyses with different datasets and approaches (I will not bother you with all the phylogenetic details. Interested readers can check the methods section of the paper). Interestingly, only one out of six phylogenetic analyses showed a sister species relationship between haplotypes H35/H36 and the Tibetan Snowcock. All the other analyses indicated that these haplotypes are more closely related to the Himalayan Snowcock. Based on these patterns, Päckert concluded that “Ding et al. (2020) did not discover previously unknown hybrid snowcocks, because haplotypes H35 and H36 just represent another deeply split mitochondrial lineage of the genetically diverse Himalayan Snowcock, T. himalayensis.”

Phylogenetic positions of the haplotypes H35/H36 based on different alignment strategies (with ClustalW on the left and with manual editing on the right) and different analyses (a = Bayesian inference and b = Maximum Likelihood). Most analyses pointed to H35/H36 as most closely related to the Himalayan Snowcock. From: Päckert (2021).

Hypervariable Regions

But what could have caused the faulty position of these haplotypes in the original study? A more detailed analysis of the mitochondrial D-loop identified the culprit, namely the hypervariable region. This section of the gene is – as the name suggests – extremely variable and can lead to errors in the alignment of genetic sequences. When Päckert removed the hypervariable region from the dataset, he uncovered a clear cluster containing the two haplotypes, the Himalayan Snowcock and the Altai Snowcock (confirming the phylogenetic results). There is thus no convincing evidence for a hybrid origin of the Altai Snowcock.

Päckert nicely summarized the lesson from this case study, namely that “results inferred from mitochondrial markers (in particular from those including hypervariable regions) require a thorough quality check.” Moreover, this example illustrates the importance of re-analyzing data and checking bold claims. That is how science progresses.

Phylogenetic networks with and without the hypervariable region clearly show the impact on the position of the haplotypes H35 and H36. From: Päckert (2021).


Ding, L., Liao, J., & Liu, N. (2020). The uplift of the Qinghai–Tibet Plateau and glacial oscillations triggered the diversification of Tetraogallus (Galliformes, Phasianidae). Ecology and Evolution, 10(3), 1722-1736.

Päckert, M. (2021). No hybrid snowcocks in the Altai—Hyper‐variable markers can be problematic for phylogenetic inference. Ecology and Evolution11(22), 16354-16364.

Featured image: Tibetan Snowcock © Donald Macauley | Wikimedia Commons

The ebb and flow of the Taiwan Strait shaped patterns of gene flow between two partridge species

Genetic analyses point to several bouts of gene flow.

The Strait of Taiwan separates the Chinese Bamboo Partridge (Bambusicola thoracicus) from the Taiwan Bamboo Partridge (B. sonorivox). It is easy to imagine that these bird species have been in contact during periods of low sea levels. And indeed, a taxonomic study from 2014 provided evidence for gene flow after their divergence, roughly 1.8 million years ago. However, these genetic analyses – using an isolation-with-migration model – only indicated that gene flow occurred, but not when. A recent paper in the journal Avian Research addressed this knowledge gap using a set of 31 nuclear loci. When did the Chinese and Taiwan Bamboo Partridge exchange genetic material?

Comparing models

The researchers compared several demographic models with different timing of gene flow. The most likely model (with a posterior probability of 0.53) pointed to early gene flow during the first 20 percent of divergence. However, a second model with late gene flow could not be rejected (posterior probability of 0.30). Together, these patterns suggest that the partridges experienced multiple bouts of gene flow. The researchers speculate that “fluctuations in the sea level of the Taiwan Strait during the early late Pleistocene may have led to changes in their distribution alternating between sympatry and allopatry.” This scenario was supported by ecological niche modelling, showing that the ranges of ancestral populations overlapped during the Last Glacial Maximum.

Three different demographic models that could explain the evolutionary history of these partridges. From: Wang et al. (2021).

Merging and diverging

The evolutionary history of the Chinese and the Taiwan Bamboo Partridge was thus shaped by multiple bouts of gene flow. As methods to detect and date gene flow events improve, we can expect to find similar scenarios in other bird species. The glacial cycles of the Pleistocene impacted the distribution of numerous species, regularly giving rise to zones of secondary contact. Many species pairs were probably subjected to cycles of merging and diverging.

These insights can help us to assess the consequences of current climate change. As species distributions change, some previously isolated populations might establish secondary contact and enter a phase of merging. These human-induced hybridization events are both a curse and a blessing. As I wrote in my review on hybridization the Anthropocene: “As humans continue to change the environment and alter species distributions, more anthropogenic hybridization events will definitely occur. This will pose challenges for the conservation of endangered species, but also provide unique opportunities for evolutionary biologists.”

Ecological niche modelling indicated overlap between both partridge species (in green) during the Last Glacial Maximum. From: Wang et al. (2021).


Hung, C. M., Hung, H. Y., Yeh, C. F., Fu, Y. Q., Chen, D., Lei, F., … & Li, S. H. (2014). Species delimitation in the Chinese bamboo partridge Bambusicola thoracica (Phasianidae; Aves). Zoologica Scripta43(6), 562-575.

Wang, P., Yeh, C., Chang, J., Yao, H., Fu, Y., Yao, C., … & Zhang, Z. (2021). Multilocus phylogeography and ecological niche modeling suggest speciation with gene flow between the two Bamboo Partridges. Avian Research12(1), 1-10.

Featured image: Chinese Bamboo Partridge (Bambusicola thoracicus) © Sun Jiao | Wikimedia Commons

Puffin population structure: There is more than meets the eye

Genetic study provides evidence for four genetic clusters.

Who doesn’t like Atlantic Puffins (Fratercula arctica)? Small black-and-white seabirds with colorful beaks that nest in rabbit-hole-like burrows. It sounds like something out of a fairytale. But these iconic birds do exist and I have had the pleasure of observing one in close proximity. On a holiday in Wales with my father and some friends, we were planning to visit a breeding colony on a nearby island. Due to the bad weather, however, the boat trip was cancelled and we were forced to change our plans. Walking along the Welsh shores, we suddenly discovered a stranded Puffin. This bird was exhausted, but still managed to bite my fathers finger. We brought it (the Puffin, not my fathers finger) to a nearby house where the surprised owner promised to deliver it at an animal shelter. We never knew if he did and what happened to this particular Puffin.

Despite their attractive looks and endangered status, little genetic work has been done on Puffins. Their current taxonomy is largely based on size differences:

  • F. a. grabae: the smallest subspecies in France, Britain, Ireland and southern Iceland
  • F. a. arctica: the intermediate subspecies in Norway, Iceland and Canada
  • F. a. naumanni: the largest subspecies in the high Arctic (e.g., Spitsbergen, Greenland, northern Canada)

A recent study in the journal Communications Biology provided a genomic perspective on this seabird species. Are these subspecies also supported by genetic data?

Four Clusters

Oliver Kersten and his colleagues compared the genetic make-up of 71 birds to detect any population structure. They could delineate four main clusters. The Puffins from Spitsbergen were clearly distinct from the other populations. And there were more subtle differences between birds from Canada, the Isle of May and multiple colonies from Iceland/Norway/Faroe Islands. Interestingly, these four genetic groups do not correspond to the three subspecies described above:

Although the genetically distinct Spitsbergen cluster coincides with the classification of morphologically large puffins in the High Arctic (F. a. naumanni), we observe gene flow from Spitsbergen into Bjørnøya, which has been considered F. a. arctica. Furthermore, the geographic divide between F. a. grabae and F. a. arctica lies farther south than previously thought, with the Faroese puffins being genetically closer to F. a. arctica than to F. a. grabae.

Moreover, the genetic population structure in the nuclear data was not observed in the mitochondrial DNA (mtDNA). The lack of clear population differentiation in mtDNA could be due to recent population expansions (but see also the seabird paradox). More detailed demographic analyses are needed to unravel the evolutionary history of the Atlantic Puffin.

The genomic data pointed to four genetic clusters (see figures c and d). This population structure was not visible in the mitochondrial DNA (figure b), suggesting a recent population expansion. From: Kersten et al. (2021).

Hybrid Population

In the quote above, you could already read that the researchers found “gene flow from Spitsbergen into Bjørnøya”. For readers unfamiliar with islands in the Arctic: Bjørnøya is a small island between Spitsbergen and Norway that houses less than 1000 breeding pairs. Although the Spitsbergen population is clearly differentiated from the other populations, it still contributes to the formation of a hybrid population on Bjørnøya. An interesting case of secondary contact that requires further investigation.

In summary, a genomic exploration of the Atlantic Puffin uncovered four distinct genetic clusters and a region of secondary contact on the small island of Bjørnøya. I wonder where the Welsh Puffin we found on holiday fits in…


Kersten, O., Star, B., Leigh, D. M., Anker-Nilssen, T., Strøm, H., Danielsen, J., … & Boessenkool, S. (2021). Complex population structure of the Atlantic puffin revealed by whole genome analyses. Communications Biology4(1), 1-12.

Featured image: Atlantic Puffin (Fratercula arctica) © Charles J. Sharp | Wikimedia Commons