Building bowers and phylogenies: Genomic study reconstructs the complex evolution of bowerbirds

Phylogenomic analyses suggest that the ability to build bowers arose twice.

In many birds species, males go to great lengths when seducing females. Think of the dazzling colors of a peacocks tail or the elaborate songs of nightingales. In one particular bird family – the Ptilonorhynchidae – males even build elaborate structures with colorful objects and plant material to capture the attention of potential female partners. These birds are better known as bowerbirds and you can check out their amazing architectural skills and captivating courtship in the video below. The resulting structures – or bowers –  are generally divided into two types: avenues and maypoles. Avenues consist of two parallel walls made of vertically placed sticks and grass stems, while maypoles are composed of sticks and other vegetation around young trees or twigs.

The exact function of these bowers remains a matter of debate, with two competing – but not mutually exclusive – hypotheses. Some ornithologists argue that these bowers are an extension of the male plumage ornaments to seduce females. Others think that the bowers provide the females with protection and prevents other males from mating with the courted female. The bowers’ function is not the only mystery about these birds. In fact, the evolutionary history of this bird family is still not entirely clear. Reconstructing the phylogeny of the bowerbirds can help scientists to better understand the evolution of bower construction and courtship. And a genomic study in the journal Systematic Biology did just that.


Nuclear DNA

In the 1990s, phylogenetic analyses of the mitochondrial gene cytochrome b provided the first molecular perspective on the evolution of bowerbirds. These studies revealed that Catbirds (genus Ailuroedus) were the sistergroup of all other bowerbirds. Catbirds are the odd one out within the family Ptilonorhynchidae: in contrast to the remaining bowerbird genera, these species are monogamous and do not build bowers. This result thus suggested that the ancestor of bowerbirds was monogamous and that the construction of bowers evolved only once.

If there is one thing we learned during the genomic revolution, it is that different genes tell different stories. Indeed, Per Ericson and his colleagues analyzed more than 12,000 nuclear loci and uncovered a different story. Catbirds are not the sistergroup of the other bowerbirds, instead they are most closely related to a group of species that build maypoles (genera Ambyornis, Prionodura) or just a courtship court (genus Scenopoeetes). The remaining bowerbirds form a monophyletic group that constructs avenue bowers.

Phylogenetic analyses of genomic data reveal three main groups: (A) avenue-building species, (B) maypole-building species, and (C) monogamous species. From: Ericson et al. (2020) Systematic Biology


Mitochondrial Mistakes

Interestingly, analyzing the complete mitochondrial genome resulted in the same phylogeny as the cytochrome b studies. What could explain the difference between mtDNA and nuclear loci? One possibility – and especially relevant for this blog – is hybridization. Perhaps an ancient hybridization event resulted in the exchange of mtDNA between the ancestors of the avenue- and maypole building birds. The authors suggest another explanation that relates to the fast evolution of mtDNA. Because mtDNA accumulated mutations faster than nuclear DNA, it becomes unreliable at estimating ancient divergences. The bowerbirds originated about 15 million years, which might be too old for stable mitochondrial analyses. This reasoning is supported by extra analyses of the mitochondrial genome. When you exclude the third codon position of mitochondrial genes – which evolves faster compared to the first two codon positions – the analyses uncover the nuclear phylogeny.

Spotted Catbird (Ailuroedus melanotis) do not build bowers. © Greg Schechter | Wikimedia Commons


Once or twice?

The new phylogeny allowed the researchers to retrace the evolution of bower-building and courtship display. Ancestral state reconstruction indicated that the ancestor of these birds did not build bowers. This result raises another question: did the ability to build bowers evolve once or twice? If this behavior arose once, it was subsequently lost in Catbirds (genus Ailuroedus) and the Tooth-billed Bowerbird (Scenopoeetes dentirostris). The latter species does not build bowers, but does prepare a courtship court to attract females.

The researchers argue for the alternative scenario: the ability to build bowers evolved twice. Once in the maypole-constructing genera Prionodura, Archboldia and Amblyornis, and another time in the avenue-building genera Ptilonorhynchus, Sericulus and Chlamydera. This scenario seems likely because both groups evolved different ways of building bowers (maypoles vs. avenues). Moreover, they note that “the relatively stable tropical and subtropical forest environment in combination with low predator pressure and rich food access (mostly fruit) are conditions that have facilitated the evolution of the extensive male displays and bower-building behavior.” Indeed, the region of the bowerbirds – Australia and New Guinea – houses another bird group with crazy courtship displays: the birds-of-paradise.

Another possible scenario that the researchers did not consider involves hybridization. Perhaps bower-building behavior evolved in one group and was consequently transferred to another group by introgressive hybridization (see my BioEssays paper for more details on this idea). Maybe the mitochondrial results do point to an ancient hybridization event and are not due to mutational saturation? This hypothesis can be tested by determining the genetic basis of bower-building, followed by phylogenetic analyses of the “bower-building genes”. If this behavior was transferred between the ancestors of these groups, these genes would cluster the bower-building genera together (similar to the mitochondrial result). To solve this mystery, we can let the birds build their bowers, while we construct some insightful evolutionary trees.

The new phylogeny enabled the researchers to reconstruct the evolution of several traits, such as plumage, social mating system, bower-building and display court. From: Ericson et al. (2020) Systematic Biology



Ericson, P. G., Irestedt, M., Nylander, J. A., Christidis, L., Joseph, L., & Qu, Y. (2020). Parallel Evolution of Bower-Building Behavior in Two Groups of Bowerbirds Suggested by Phylogenomics. Systematic Biology, 69(5): 820-829.

Featured image: Satin Bowerbird (Ptilonorhynchus violaceus) © Joseph C. Boone | Wikimedia Commons


This paper has been added to the Ptilonorhynchidae page.