Look into my eyes: Unraveling the genetic basis of iris color in domestic pigeons

A nonsense mutation in one gene underlies the “pearl” phenotype.

All roads lead to Rome. This saying was nicely illustrated by three research groups that independently discovered the genetic basis of the “pearl” eye color in the domestic pigeon (Columba livia). When you take a closer look at pigeons, you might notice three main eye colors: orange, pearl (white), and bull (dark brown). One of the three papers – by Emily Maclary and her colleagues – provided a nice summary of these three eye colors.

Orange iris color is the ancestral state, and “orange” eyes in actuality range in shades from yellow to red, depending on the density of blood vessels in the eye. The pearl iris color is white, with tinges of pink and red from blood vessels. Lastly, the bull iris color is named based on its similarity in color to dark bovine eyes, and ranges from dark brown to almost black.

In this blog post, I will focus on the genetic basis of the pearl eye color. Interested readers can check this Molecular Biology and Evolution study for more information on the genetics of the bull eye color.

One Candidate Gene

Using different approaches – from genome-wide association studies to laboratory crosses – the three research groups all zoomed in on the gene SLC2A11B (solute carrier family 2, facilitative glucose transporter, member 11b). Detailed analyses of this gene revealed a nonsense mutation that results in a premature stop codon. In their PLoS Genetics study, Si Si and co-workers noted that “the W49X mutation leads to the truncation of approximately 90% of the amino acids of SLC2A11B and is predicted to cause a total loss of function in the protein.”

Comparing the genomes of domestic pigeons with pearl-eye and wild-type colors revealed a clear candidate region that contained the gene SLC2A11B. From: Andrade et al. (2021).

Pteridine Pigments

In another PLoS Genetics study, Pedro Andrade and his colleagues quantified the expression of the candidate gene. They found that the gene SLC2A11B was down-regulated in pigeons with the pearl eye color. This expression pattern suggests that the nonsense mutation creates an aberrant RNA-molecule that is broken down by the cell. The product of SLC2A11B might thus be non-functional.

A functional SLC2A11B gene would be involved in the production of pteridine pigments which give the wild-type eye its orange color. No functional SLC2A11B gene means no pteridine pigments, resulting in a pearl iris color.

Expression data indicated that the SLC2A11B gene is down-regulated in pigeons with the pearl eye color. From: Andrade et al. (2021).

Artificial Selection

Finally, we can have a look at the evolutionary history of this mutation. Si Si and co-workers performed a phylogenetic analysis on 139 domestic pigeons (35 fancy pigeons, two feral pigeons, and 102 racing pigeons) and one Hill Pigeon (Columba rupestris, as an outgroup). The resulting evolutionary tree showed that the nonsense mutation arose approximately 5,400 years ago, coinciding with the possible start of pigeon domestication in the Fertile Crescent. The researchers also detected strong signals of positive selection pointing to artificial selection for this trait.

If you have read the Origin of Species, you know that Darwin used the case of artificial selection in pigeons as a starting point for his long argument in favor of evolution by natural selection. These studies would definitely have been an eye-opener for him.

Phylogenetic analyses revealed that the nonsense mutation (indicated in the orange branches) originated once about 5400 years ago. From: Si et al. (2021).

References

Andrade, P., Gazda, M. A., Araújo, P. M., Afonso, S., Rasmussen, J. A., Marques, C. I., … & Carneiro, M. (2021). Molecular parallelisms between pigmentation in the avian iris and the integument of ectothermic vertebrates. PLoS Genetics, 17(2), e1009404.

Maclary, E. T., Phillips, B., Wauer, R., Boer, E. F., Bruders, R., Gilvarry, T., … & Shapiro, M. D. (2021). Two genomic loci control three eye colors in the domestic pigeon (Columba livia). Molecular Biology and Evolution, 38(12), 5376-5390.

Si, S., Xu, X., Zhuang, Y., Gao, X., Zhang, H., Zou, Z., & Luo, S. J. (2021). The genetics and evolution of eye color in domestic pigeons (Columba livia). PLoS Genetics, 17(8), e1009770.

Featured image: Common Pigeon (Columba livia) © Satdeep Gill | Wikimedia Commons

The evolution of complex behaviors in cranes

The common ancestor of cranes already had an impressive behavioral repertoire.

Reconstructing phylogenetic trees remains an important aspect of evolutionary biology. By knowing how different species are related to each other, we can test hypotheses about their evolutionary history. In the past, researchers would rely on morphological data to build phylogenetic trees. However, these morphological datasets could be misleading if distantly related species evolved similar traits (a phenomenon known as homoplasy). The advent of genetic sequencing revolutionized the study of phylogenetic relationships, although this type of data came with its own challenges.

But what about behavioral data? In the 1993, Alan de Queiroz and Peter Wimberger nicely characterized the attitude of evolutionary biologists towards the phylogenetic usefulness of behavioral traits.

It is widely believed that behavior is more evolutionarily labile and/or more difficult to characterize than morphology, and thus that behavioral characters are not as useful as morphological characters for estimating phylogenetic relationships.

This prejudice was proven wrong by the meta-analysis of de Queiroz and Wimberger. They analyzed 22 data sets with both morphological and behavioral characters and found no significant differences between their phylogenetic consistency. Behavioral data can thus be used in phylogenetic analyses. But how reliable are the results?

Crane Phylogeny

In a recent Avian Research study, Nela Nováková and Jan Robovský reconstructed the evolutionary history of cranes (family Gruidae) based on an extensive dataset of 107 behavioral traits. However, only 28 traits turned out to be phylogenetically informative, because most traits were shared by all crane species, or were unique to one species. The reduced dataset could not confidently resolve the phylogeny of the cranes, although some subgroups did emerge.

From these relationships only several subgroups are concordant with the topology based on mitochondrial genome – the subgroup comprising Balearica species, Bugeranus with Antropoides, and all species of the genus Grus (with G. japonensis sister to other Grus species) respectively.

A phylogenetic analysis based on 48 morphological traits provided better results. In this case, morphological data did perform better than behavioral data.

Ancestral Species

Next, the researchers combined the behavioral data with a phylogenetic analysis of mitochondrial genomes. This approach allowed them to reconstruct the behavioral repertoire of ancestral crane species. Interestingly, it seems that the cranes lost more traits than they gained, suggesting that the common ancestor of all cranes had an elaborate behavioral repertoire. This pattern could be explained by the idea that losing a trait is easier than evolving a new one.

Unfortunately, because behavior does not fossilize, we cannot be completely sure what the behaviors of these ancestral species looked like. Wouldn’t it be great if we could travel back in time and observe the courtship behavior of these extinct crane species? Who knows what peculiar dance moves they performed.

Crane phylogeny based on complete mtDNA with mapped homologous (black) and homoplasious (white) changes in behavioral traits. From: Nováková & Robovský (2021).

References

de Queiroz, A., & Wimberger, P. H. (1993). The usefulness of behavior for phylogeny estimation: levels of homoplasy in behavioral and morphological characters. Evolution, 47(1), 46-60.

Nováková, N., & Robovský, J. (2021). Behaviour of cranes (family Gruidae) mirrors their phylogenetic relationships. Avian Research, 12(1), 1-11.

Featured image: Sandhill Crane (Antigone canadensis) © Patrick Myers (Great Sand Dunes National Park and Preserve) | Wikimedia Commons

Are the intermediate migration routes of hybrids ecologically inferior?

Ecological niche modelling provides support for a common assumption.

Simple experiments can provide crucial insights. In the 1990s, Andreas Helbig studied the migratory behavior of European Blackcaps (Sylvia atricapilla) in the lab. These small passerines orient their migration either southwest or southeast. Hybrids between individuals that use different migratory strategies showed an intermediate migration route, namely directly to the south.

Contact zones between populations that use different migratory routes are known as migratory divides. I have covered several examples on this blog, such as Barn Swallows (Hirundo rustica) and Red-necked Phalaropes (Phalaropus lobatus). Most studies on migratory divides assume that hybrids have reduced fitness because their intermediate migration takes them over less suitable habitat. An intuitive assumption that is rarely assessed. A recent study in the journal Global Ecology and Biogeography put this common assumption to the test.

Ecological Modelling

Hannah Justen and her colleagues focused on the well-studied migratory divide between two subspecies of the Swainson’s Thrush (Catharus ustulatus). Previous work with geolocators revealed that “coastal birds [subspecies swainsoni] migrate along the west coast of North America between British Columbia and Mexico, Guatemala or Honduras, whereas inland birds [subspecies ustulatus] migrate along more eastern routes between British Columbia and Colombia or Venezuela.” Hybrids show a range of migratory strategies, some birds follow the parental routes while others migrate along intermediate trajectories.

But are intermediate routes really ecologically inferior? The researchers tested this assumption with ecological niche modelling and measures of landscape connectivity. Together, these analyses indicated that “the intermediate area between parental migratory ranges is generally less suitable for migration and more costly to move through.” These findings provide convincing support for the common assumption that some hybrids follow suboptimal migration routes.

Comparison of the predicted suitability of sites within the parental and intermediate migratory ranges of Swainson’s Thrushes for spring and autumn migration. From: Justen et al. (2021).

Carry-over Effects

But the story does not end here. A closer look at the results revealed that some intermediate migration routes contain pockets of high habitat suitability. It is thus possible that some hybrid thrushes can successfully migrate between their breeding and wintering ranges. However, these birds might suffer from carry-over effects in which body condition after migration impacts reproductive success. The logical next step is to directly quantify the fitness of hybrids with intermediate migration routes. More work to support a seemingly simple assumption.

References

Justen, H., Lee‐Yaw, J. A., & Delmore, K. E. (2021). Reduced habitat suitability and landscape connectivity in a songbird migratory divide. Global Ecology and Biogeography, 30(10), 2043-2056.

Featured image: Swainson’s Thrush (Catharus ustulatus) © Wildreturn | Wikimedia Commons

No evidence for hybrid speciation in the Ashy-throated Parrotbill

Admixed individuals do not necessarily point to a hybrid origin.

Hybrid speciation – the origin of new species through hybridization – seems to be a rare phenomenon in birds. In my 2018 review paper on this topic, I could only find convincing evidence for four putative hybrid bird species. In recent years, several additional species have been proposed to be of hybrid origin, such as Salvin’s Prion (Pachyptila salvini), Steller’s Eider (Polysticta stelleri), and Altai Snowcock (Tetraogallus altaicus). The evidence supporting these hybrid species is mixed (as you can read in the linked blog posts). Recently, another putative hybrid species was added to the list: the Ashy-throated Parrotbill (Paradoxornis alphonsiana).

In their Molecular Phylogenetics and Evolution study, Chuanyin Dai and Ping Feng presented genetic analyses of the Paradoxornis webbianus species complex across East Asia. Based on a dataset of one mitochondrial and five nuclear markers, they concluded that “P. alphonsianus was likely the result of hybridization between P. webbianus and P. a. ganluoensis.” Martin Päckert reevaluated the genetic evidence underlying this bold statement. He published his findings in the Journal of Avian Biology.

Admixed Individuals

As you might have guessed from the title of this blog post, Päckert does not agree with the hybrid speciation scenario. His main point of criticism relates to the simplistic reasoning of Dai and Feng. They found several admixed individuals in a genetic clustering analysis (using the software STRUCTURE) and immediately jumped to the hybrid species conclusion. As Päckert nicely put it: “the mere presence of admixed individuals in a population is insufficient evidence for hybrid speciation.”

Indeed, more analyses are needed to confidently discriminate between hybrid speciation and other processes, such as recurrent introgressive hybridization or incomplete lineage sorting. In another blog post, I proposed several approaches to test for a hybrid speciation scenario, such as coalescent modelling.

The genetic clustering analyses of Dai and Feng (2023) uncovered several admixed individuals (with blue and yellow ancestry). However, these patterns alone are insufficient to support a hybrid speciation scenario. From: Päckert (2023).

Genomic Power

Interestingly, Dai and Feng did not consider the possibility of incomplete lineage sorting (ILS). The genetic patterns in their STRUCTURE plot could also be the outcome of this process (see this blog post for more details). In fact, a previous genetic study on parrotbills reported that two nuclear markers (IRF and TGFB) were impacted by ILS. This information suggests that the set of molecular markers was not powerful enough to confidently discriminate between hybridization and ILS. Genomic analyses are needed to reassess the potential patterns of gene flow within the Paradoxornis webbianus species complex.

In his critique, Päckert provided a nice example to illustrate the difference in explanatory power between a handful of molecular markers and genome-wide data.

That Dai and Feng’s (2023) set of five intron markers was a suboptimal choice for their parrotbill study, is reflected by the strongly contrasting divergence and admixture patterns inferred from different sets of markers across a Chinese contact zone of bush tits (Aegithalos fuliginosus and Ae. bonvaloti) in Sichuan. A first study by Wang et al. (2014) relied on six nuclear loci only and could not detect any signal of gene flow between the two species. In contrast, a successive study on approximately 70 000 genome-wide SNPs could detect strong divergence in allopatric populations but also a signal of admixture in the contact zone of these two bush tit species (Zhang et al. 2017). 

In the end, the genetic evidence presented by Dai and Feng is insufficient to support a hybrid speciation scenario for the Ashy-throated Parrotbill. A good occasion to use on of my favorite Carl Sagan quotes: “extraordinary claims require extraordinary evidence”.

References

Dai, C., & Feng, P. (2023). Multiple concordant cytonuclear divergences and potential hybrid speciation within a species complex in Asia. Molecular Phylogenetics and Evolution, 180, 107709.

Päckert, M. (2023). No evidence of a hybrid origin of the ashy‐throated parrotbill Sinosuthora alphonsiana. Journal of Avian Biology, 2023(11-12), e03146.

Qu, Y., Zhang, R., Quan, Q., Song, G., Li, S. H., & Lei, F. (2012). Incomplete lineage sorting or secondary admixture: disentangling historical divergence from recent gene flow in the Vinous‐throated parrotbill (Paradoxornis webbianus). Molecular Ecology, 21(24), 6117-6133.

Featured image: Vinous-throated Parrotbill (Paradoxornis webbianus) © Alnus | Wikimedia Commons

Note on taxonomy: The genus name for these parrotbills differed between publications. Dai and Feng (2023) referred to Paradoxornis whereas Päckert (2023) used Sinosuthora. To avoid confusion, I followed Dai and Feng (2023) because I quoted from their paper at the beginning of the blog post.