Genomics of speciation in a great speciator (Aves: Zosterops) reveals the roles of both natural and sexual selection

This study utilizes genomic data from African and Indian Ocean white-eyes to demonstrate that reproductive isolation in this "Great Speciator" lineage is driven by a combination of natural and sexual selection, with the Z chromosome playing a central role in divergence, particularly when ecological differences are involved.

The Gist

Imagine a tiny, energetic bird called the White-eye living on a volcanic island in the Indian Ocean. Scientists call this bird a "Great Speciator," which is a fancy way of saying it's a master at splitting into new, distinct groups very quickly.

This paper is like a genomic detective story. The researchers wanted to figure out how these birds are turning into different species so fast. They looked at the birds' DNA (their biological instruction manual) to find the specific "switches" that are causing them to separate.

Here is the story of their findings, broken down simply:

1. The Setting: A Bird Family with a Split Personality

On the island of Réunion, these White-eyes live in two very different neighborhoods:

• The Lowlands: Three groups of birds live here. They look slightly different (different shades of brown and grey on their heads) but live in similar weather. They are separated by rivers and lava flows, not by climate.
• The Highlands: A fourth group lives high up in the mountains (above 1,400 meters). It's colder, the air is thinner, and the plants are different. These birds are bigger and have different colors.

Even though they live on the same small island, these groups don't mix much. They have formed "neighborhoods" that rarely cross the street. The scientists wanted to know: What is the genetic glue holding these groups apart?

2. The Investigation: Three Different Detective Tools

To solve the mystery, the team used three different "flashlights" to scan the birds' DNA:

• Flashlight 1: The "Hotspot" Map. They looked for areas in the DNA where the lowland birds and highland birds were very different from each other. But here's the trick: DNA has "traffic jams" (areas where genes don't swap easily). Sometimes, differences happen just because of the traffic jam, not because the birds are evolving. The scientists used a map of "traffic flow" (recombination rates) to ignore the jams and only look at the differences that actually matter.
• Flashlight 2: The "Sweep" Detector. They looked for signs that a specific gene had been "swept" through a population because it was super useful. Imagine a wave of a new, better phone model sweeping through a city; the old models disappear. They looked for these waves in the birds' DNA.
• Flashlight 3: The "Gatekeeper" Calculator. They used a computer model to simulate how the birds moved and mixed over time. This helped them identify which specific genes act like "gates" that stop the birds from breeding with each other.

3. The Big Discovery: Two Different Reasons for Splitting

The study found that the birds are separating for two different reasons, depending on which group you look at:

A. The Lowland Birds: The "Fashion & Music" Split
The three lowland groups live in similar weather, so they aren't fighting over survival. Instead, they are splitting because of Sexual Selection.

• The Analogy: Think of them like different music subcultures. One group likes "Brown Head" fashion and sings a specific song. Another group likes "Grey Head" fashion and a different tune.
• The Genes: The DNA differences were mostly found on the Z chromosome (the bird equivalent of the X/Y sex chromosomes). This is where genes for song and feather color live. The birds are essentially saying, "I only want to mate with someone who looks like me and sings like me."

B. The Highland Birds: The "Survival" Split
The highland birds are separated from the lowland birds because they live in a totally different world.

• The Analogy: Imagine a group of people moving from a beach town to the top of Mount Everest. They need new gear to survive the cold and thin air.
• The Genes: Here, the differences were found all over the DNA (not just the Z chromosome). The genes involved are for breathing in thin air, metabolism, immune systems (fighting different mountain germs), and beak shape (to eat different mountain food). This is Natural Selection at work: "I need to change my body to survive here."

4. The "Great Speciator" Paradox

Usually, scientists think speciation (becoming a new species) takes millions of years. But these White-eyes are doing it in the blink of an evolutionary eye (less than 500,000 years).

The paper concludes that this bird is a "Great Speciator" because it uses a double-engine approach:

1. Sexual Selection (fashion and music) keeps the lowland groups apart.
2. Natural Selection (survival gear) keeps the highland group apart from the rest.

The Takeaway

This study is like watching a magic trick in real-time. It shows us that nature doesn't need a million years to create new species. Sometimes, if you have a bird that loves to sing, a bird that changes its color, and a bird that can adapt to the mountains, you can get a whole new family tree sprouting up right on your doorstep.

The "Great Speciator" isn't just a bird; it's a master architect, using the tools of beauty and survival to build walls between its own children, turning one species into many in record time.

Technical Summary

1. Problem Statement and Objective

The study addresses the challenge of distinguishing between background selection (neutral processes linked to recombination landscapes) and divergent selection (adaptive processes driving reproductive isolation) in genomic landscapes of differentiation. While "islands of speciation" (genomic regions of high differentiation) are often attributed to divergent selection, they can also arise from reduced recombination rates where background selection is strong.

The authors utilize the Reunion Grey White-eye (Zosterops borbonicus), a "Great Speciator" endemic to the island of Réunion, as a natural experimental model. This species complex comprises:

• Three lowland forms (LBHB, GHB, BNB) separated by physical barriers (rivers/lava) but sharing similar ecological niches, likely driven by sexual selection (plumage/song).
• One highland form (HIGH) occupying a distinct elevational zone (>1,400m), likely driven by natural selection (altitude adaptation).

The objective was to identify candidate barrier loci responsible for reproductive isolation across this speciation continuum, disentangling the effects of natural vs. sexual selection, and accounting for recombination rate variation.

2. Methodology

The researchers employed a multi-faceted genomic approach using whole-genome resequencing data from 34 individuals (including Z. borbonicus forms and outgroup species: Z. mauritianus, Z. olivaceus, Z. pallidus, and Z. virens).

Key Analytical Steps:

1. Data Processing: Reads were mapped to a high-quality Z. borbonicus reference genome. Approximately 13 million SNPs were called for inter-species comparisons and 7 million for intra-species (form) comparisons.
2. Differentiation Landscapes ($F_{ST}$): Pairwise $F_{ST}$ was calculated in 50-kb windows. The distribution of $F_{ST}$ was analyzed to assess skewness and accumulation of differentiation across the speciation continuum.
3. Recombination Correction: To distinguish selection from background selection, the authors:
   • Correlated $F_{ST}$ with recombination rates (extrapolated from the Zebra Finch map due to high conservation in birds).
   • Identified "candidate barrier loci" as regions with high $F_{ST}$ (top 0.5%) AND moderate-to-high recombination (>1 cM). This filters out regions where high differentiation is merely an artifact of low recombination.
   • Used comparative genomics with outgroup species to identify shared high-$F_{ST}$ regions (likely background selection) versus lineage-specific outliers (likely adaptive).
4. Selective Sweep Detection: Used Sweepfinder2 to detect Composite Likelihood Ratio (CLR) signals of recent selective sweeps in each form, incorporating recombination maps to reduce false positives.
5. Demographic Inference (ABC): Used DILS (Demographic Inferences with Linked Selection) with Approximate Bayesian Computation to model gene flow. This identified loci with significantly reduced migration rates ("barrier loci") specifically in the Highland vs. Lowland comparison, where heterogeneous migration was detected.
6. Functional Annotation: Genes within candidate regions were annotated using Zebra Finch synteny and manually curated literature searches to infer biological functions.

3. Key Results

A. Genomic Landscapes and Recombination

• Heterogeneity: Genomic differentiation is highly heterogeneous. As divergence increases, $F_{ST}$ accumulates, and the distribution becomes less skewed.
• Z Chromosome Effect: In recently diverged lowland forms, the Z chromosome showed markedly lower skewness in $F_{ST}$ distribution compared to autosomes, suggesting it plays a disproportionate role in early reproductive isolation.
• Recombination Correlation: A negative correlation exists between $F_{ST}$ and recombination rate. However, the study successfully identified specific regions with high $F_{ST}$ and high recombination, confirming they are likely targets of positive selection rather than background selection artifacts.

B. Candidate Barrier Loci by Selection Regime

• Lowland vs. Lowland (Sexual Selection):
  • Candidate loci were predominantly concentrated on the Z chromosome.
  • Genes involved song behavior (e.g., KCNS2, KCNV2), plumage pigmentation (e.g., AIM1, ASCC3), and neurobehavioral traits.
  • Selective sweeps were restricted to the Z chromosome in lowland forms.
• Highland vs. Lowland (Natural Selection):
  • Candidate loci were distributed across both autosomes and the Z chromosome.
  • Genes were enriched for altitude adaptation (hypoxia response: MRPL33, EPRS, JAK2), metabolic regulation, immune function, and morphology (e.g., VPS13B linked to bill shape).
  • Selective sweeps in highland forms were distinct between the two mountain ranges (North vs. South), suggesting local adaptation even within the highland form.

C. Demographic Inference

• Heterogeneous gene flow (barrier loci) was only detected in the Highland vs. Lowland comparison.
• No significant barriers were detected between lowland forms, likely due to low overall differentiation and the power limits of the method, though narrow hybrid zones exist.
• Barrier loci in the Highland/Lowland comparison contained genes related to hypoxia, immunity, and pigmentation, supporting a mix of pre- and post-mating isolation mechanisms.

4. Key Contributions

1. Methodological Rigor: The study demonstrates the necessity of integrating recombination landscapes and comparative genomics (using outgroups) to filter out background selection signals. By focusing on high-$F_{ST}$ regions in high-recombination areas, the authors isolated genuine adaptive signals.
2. Disentangling Selection Types: The research provides empirical evidence that sexual selection (driving lowland divergence) primarily targets the Z chromosome, while natural selection (driving altitudinal divergence) targets both autosomes and the Z chromosome.
3. Gene Function Identification: The study links specific genomic regions to functional traits critical for speciation:
   • Song/Pigmentation: Drivers of assortative mating in lowlands.
   • Hypoxia/Metabolism: Drivers of ecological adaptation in highlands.
4. The "Great Speciator" Paradox: It offers a genomic explanation for how Zosterops species can diversify rapidly on small spatial scales, showing that reproductive barriers can evolve quickly through a combination of sexual and natural selection.

5. Significance

This paper significantly advances the field of speciation genomics by moving beyond simple "island of differentiation" identification. It validates that the genomic architecture of divergence is dynamic and context-dependent:

• Early-stage divergence (lowland forms) is characterized by Z-chromosome linkage, potentially due to the "large X/Z effect" in reproductive isolation.
• Ecological divergence (highland forms) involves a broader genomic footprint, reflecting polygenic adaptation to environmental gradients.

The findings highlight that reproductive isolation in this system is driven by a synergy of natural and sexual selection, acting on different parts of the genome depending on the ecological context. This provides a robust framework for understanding rapid speciation in other "great speciator" lineages and underscores the importance of considering recombination rates when interpreting genomic scans for selection.