Recombining ancestral variation to generate phenotypic novelty

Introgression and selection of ancestral genetic variation underlies diversification in a remarkable radiation of birds.

Like Comment

The paper in Nature Ecology & Evolution is here:

Understanding the origins of biodiversity is a central goal of evolutionary biology. We often assume that novel mutations are a primary source of new phenotypes, and thus play a key role in generating diversity. In contrast, existing genetic variation is more commonly thought to facilitate adaptive radiations that involve convergence, providing a mechanism for multiple species or populations to evolve similar phenotypes. Classical examples of this are convergent phenotypes in Heliconius butterflies and three-spined sticklebacks that depend on ancestral alleles at functional loci. However, standing genetic variation may also contribute to the evolution of novel phenotypes if different combinations of alleles are generated through differential sorting and/or introgression.

Comparing the genomes of closely related species allows us to identify loci that have presumably been targets of divergent selection, but pairwise comparisons provide little insight into the origins of adaptive alleles at these loci. For this, it is necessary to examine patterns of genomic variation in entire groups of closely related species.

A map showing phenotypic variation and geographic distributions of 11 of the 13 species in the Lonchura radiation.

My dissertation research addressed this question in a remarkable radiation of birds, the Lonchura munias of Papua New Guinea and Australia. Based on mitochondrial DNA sequences, primarily from old museum specimens, my graduate advisor and coauthor had known for many years that a clade of 13 munia species might represent one of the most recent and rapid radiations ever described in birds. Descended from a recent common ancestor, the munia species are remarkably similar genetically but show substantial variation in plumage color and pattern as well as in bill size and color. Intriguingly, these birds also have an unusual geographic distribution for recently evolved species, with two or even three of the 13 species co-occurring in at least six different regions in New Guinea and Australia. Despite this broad range overlap, reports of hybrid individuals are rare.

Chestnut-breasted Munias (Lonchura castaneothorax) near Alotau in Milne Bay Province, Papua New Guinea. Photo by Jacob Berv.

As most of these species had never been the subject of intensive scientific study and very few tissue samples from wild birds were available, a thorough investigation of the group would require extensive fieldwork in Papua New Guinea, a task I undertook as a graduate student. Over the course of two years, I traveled with a series of capable assistants and guides to Western Australia and six different provinces in Papua New Guinea to collect samples for genetic analysis. The munias are seed-eating birds that have thrived on introduced grasses and agricultural crops, and my field sites were often near human habitation, though this ranged from large towns to remote villages that had to be reached by boat or by Cessna. I sampled 11 species from 18 populations, including populations from each region where multiple munia species from this radiation co-occur.   

The author removing a bird from a mist net near Owa in Western Province, Papua New Guinea. Photo by Michael Sorenson.

We examined genomic variation in these populations using double-digest RAD sequencing (ddRAD-seq) and whole genome sequencing (WGS). The ddRAD-seq data revealed evidence of extensive genome-wide introgression between sympatric pairs of munia species. In most cases, sympatric populations of different species were more similar to one another genome-wide than allopatric populations of the same species. However, whole genome sequencing comparisons revealed several regions that exhibit elevated divergence in multiple pairs of sympatric populations. Genetic variants at these regions have discordant phylogenetic histories, suggesting that the munias have acquired different sets of alleles via shared ancestry and hybridization.

Unique combinations of alleles at these divergent loci, which include several well-characterized color genes, are likely responsible for the distinct phenotypes of the different munia species. We use the analogy of a Mr. Potato Head toy, in which evolution has created unique phenotypes by mixing and matching ancestral variants in different ways. This suggests that ancestral genetic variation may be important not only in examples of convergent evolution, but also as a source of genetic raw material for generating phenotypic novelty and rapid diversification.

Katherine Faust Stryjewski

Postdoctoral Fellow, Harvard University