Darwin's Finches: The Evolution Experiment That Changed Biology

When the HMS Beagle reached the Galápagos Islands in September 1835, Charles Darwin collected a handful of small brown birds without bothering to record which island each came from. Those birds, now called Darwin's finches, turned out to be one of the best natural experiments in biology. Their beaks range from thick and crushing to slender and probing, and the differences track what each species eats.

The story has two halves. The first is Darwin's, and it is messier than the legend. The second belongs to Peter and Rosemary Grant, who spent more than four decades measuring finches on a tiny island and caught natural selection in the act. This article follows both, then looks at the genes behind the beaks and at what the finches cannot tell us.

What Darwin actually did in the Galápagos

The Beagle arrived at the archipelago on September 15, 1835, and sailed for Tahiti on October 20, according to the timeline summarized in Wikipedia's account of the second voyage of HMS Beagle. Darwin was 26. He spent part of that time ashore on James Island with the ship's surgeon, Benjamin Bynoe, while the Beagle went for fresh water. He paid the finches little attention, and he did not label them by island.

The mockingbirds were a different matter. Darwin did label those by island, and he noticed that each island seemed to have its own form. Back aboard ship he wrote that each variety was constant on its own island. That observation, not the finches, was what first nagged at him.

The finches only made sense later. In 1837 the ornithologist John Gould examined Darwin's specimens at the Zoological Society of London and concluded that birds Darwin had taken for blackbirds, "gross-bills," and finches formed a single new group containing 12 species. They were not a scatter of unrelated birds that happened to look alike. They were close relatives with sharply different beaks.

Because Darwin had not recorded the islands for his finch specimens, he had to borrow labeled birds collected by Captain Robert FitzRoy and other crew members to work out which species lived where. That patchwork of sources is the origin of the persistent myth that Darwin stood on a beach, looked at beaks, and saw evolution. He did not. The insight accumulated over the following two decades, and it leaned on many lines of evidence: barnacles, pigeons, fossils, and geography.

For a fuller account of how that idea developed, see Charles Darwin: The Idea That Took 20 Years to Publish and the biography at Charles Darwin: The Reluctant Revolutionary.

The idea the finches illustrate

Darwin's mechanism has four parts, and each can be checked against the birds.

1. Variation. Individuals in a population differ in measurable traits, such as beak depth.
2. Heritability. Some of that variation is passed from parents to offspring. Large-beaked parents tend to have large-beaked chicks.
3. Differential survival and reproduction. In a given environment, some variants leave more offspring than others.
4. Accumulation. Over generations, the favored variants become more common.

Think of it as a sieve rather than a sculptor. The environment does not shape a beak to a plan. It simply lets some beaks through and not others, and the population that comes out the other side looks different from the one that went in.

Applied to the finches, the founding story runs like this. A small group of finches from the South American mainland reached the islands. Wikipedia's summary of the research places that arrival more than a million years ago. With few competitors and varied food, their descendants split into species that specialize on large hard seeds, small seeds, cactus, insects, or buds. Biologists call that kind of rapid branching into many ecological roles an adaptive radiation (the diversification of one lineage into many forms that exploit different resources).

Darwin published On the Origin of Species in 1859, and he lacked a working theory of inheritance. Gregor Mendel's pea experiments, which supplied the missing logic, appeared in 1866 and were largely ignored until around 1900. You can read about them in Mendel's Pea Plant Experiments: How Genetics Was Born. Darwin also could not watch selection operate. A century later, two researchers could.

Daphne Major: a laboratory the size of a few city blocks

Daphne Major is a small, uninhabited volcanic island in the Galápagos. It has few plant species, no permanent human presence, and a finch population small enough to count. Those features make it ideal. The Grants and their students could catch, band, and measure essentially every bird, then follow each one's fate.

The scale of the project is hard to overstate. Wikipedia's finch article describes the research as running for more than 40 years, and the Grants' long-term datasets span multiple generations of two main study species, the medium ground finch (Geospiza fortis) and the cactus finch (G. scandens). A paper in Science in 2002 summarized 30 years of data, from 1972 to 2001, for those two species. Grant and Grant's 2002 paper, "Unpredictable evolution in a 30-year study of Darwin's finches," is the best single entry point to the long record.

The measurements themselves are modest and repeatable: body mass, wing length, beak length, beak depth, and beak width, taken with calipers. What made the dataset powerful was that the same people did it year after year, and that they paired the measurements with survival records, breeding records, and rainfall.

The 1977 drought: selection you can measure

Rain on Daphne Major is erratic. In 1977 almost none fell. The teaching-grade summary hosted by the Open University notes that many plants produced few or no seeds, and that small seeds became exceedingly rare while large seeds with thick husks were still available. Rainfall that year was close to nothing.

The effect on the medium ground finch was severe. The Open University module reports that the population fell from about 1,400 birds to a few hundred in just over two years. Counts vary with the start and end dates used, so the safest statement is that the great majority of birds died.

The deaths were not random. Here is the comparison the Grants reported in their 1981 paper, as summarized by the Open University:

Group	Average beak depth
Medium ground finches before the drought	9.2 mm
Breeding survivors after the drought	9.9 mm
Chicks hatched the following year (1978)	9.7 mm

Large-beaked birds could crack the thick-husked seeds that remained. Small-beaked birds could not, and starved. The survivors averaged about 0.7 mm deeper in the beak, which sounds trivial until you remember that beak depth in this species is only about 9 mm. That is a shift of roughly 7 to 8 percent in the survivors. The next generation, which inherited the beak size of its parents, came out about 0.5 mm deeper than the pre-drought average.

Boag and Grant's 1981 paper in Science, "Intense natural selection in a population of Darwin's finches (Geospizinae) in the Galápagos," reported that survival was nonrandom and that large-beaked birds, particularly males, survived best. The authors described the selection as intense.

A population lost most of its members in under two years, and the survivors passed a measurable change in beak depth to their chicks. That is natural selection with a ruler and a notebook.

Two features make this a clean demonstration. First, beak depth is heritable, so the change carried into the next generation rather than vanishing when conditions improved for individuals. Second, the Grants had measured the birds before the drought, so they did not have to infer the starting condition from a later snapshot.

Why evolution did not simply keep going

A reasonable guess is that after 1977 the finches kept marching toward bigger and bigger beaks. They did not. The 2002 paper's central finding is that short-term evolution can be predicted from knowledge of selection and inheritance, but the long-term trajectory cannot, because environments fluctuate unpredictably.

Over the 30 years covered by the study, both species changed several times in body size and in two beak traits. Selection varied in form: it was sometimes consistent in direction and sometimes oscillating, sometimes sudden and sometimes gradual. Neither species ended the period in a state anyone could have predicted at the start.

That is the opposite of the popular picture of evolution as a one-way ladder. When wetter years returned and small seeds became plentiful again, large beaks lost their advantage. Small-beaked birds, which were better at handling small seeds, did comparatively better. Selection tracked the environment, and the environment kept changing.

The practical lesson is that "fitness" is not a property of a beak. It is a property of a beak in a particular year on a particular island. A deep beak is an asset in a drought and a liability, or at least neutral, when soft seeds are abundant.

Hybrids and the finch that became its own lineage

For most of the twentieth century, biologists treated hybrids between species as evolutionary dead ends. The finches complicated that view.

The Grants recorded that hybridization on Daphne Major was rare but recurring, and the 2002 paper reports that it raised the variation in beak traits in the cactus finch and changed its beak shape. Hybrids were not a curiosity. They were a source of new combinations of traits that selection could act on.

The most striking case began with a single bird. In 1981, a male arrived on Daphne Major that was unusual in song and noticeably larger in body and beak than the island's resident species. Genome sequencing later showed he was a male Geospiza conirostris from Española, an island more than 100 km away. He bred with a resident medium ground finch, and his descendants became known as the "Big Bird" lineage.

Lamichhaney and colleagues' 2018 paper in Science, "Rapid hybrid speciation in Darwin's finches," reports that the lineage bred only within itself from the second generation onward. It survived despite intense inbreeding, and its bills were outside the range of either parent species. The authors estimate that reproductive isolation, which they note usually takes hundreds of generations to develop, was established in about three generations.

Two cautions belong here. The Big Bird lineage is a single, small population, and its long-term survival is uncertain. And popular accounts sometimes blur this episode with others. The correct sequence is a 1981 immigrant, a hybrid pairing, and a published genetic analysis in 2017 and 2018. It is a strong example of how a new lineage can arise quickly. It does not show that every species forms that way.

The genes behind the beaks

Measuring beaks shows that selection works. It does not show what changed inside the birds. Four sets of results, listed here in the order they were published, begin to answer that.

Year	Gene or region	What the study found	Source
2004	Bmp4	Expression in the upper-beak mesenchyme correlated with deep, broad beaks across Geospiza species. Misexpression in chicken embryos produced beak changes resembling the large ground finch.	Abzhanov et al., Science
2006	Calmodulin pathway	Higher expression in cactus finches, which have long, pointed beaks. Upregulating the pathway in chick embryos lengthened the upper beak.	Abzhanov et al., Nature
2015	ALX1	A 240-kilobase haplotype containing this craniofacial gene was strongly associated with beak shape across species and within the medium ground finch.	Lamichhaney et al., Nature
2016	HMGA2 region	A locus associated with beak size appeared to have facilitated character displacement during a drought.	Lamichhaney et al., Science

The 2015 study re-sequenced the whole genomes of 120 individuals covering all Darwin's finch species and two close relatives. It found that the genetic family tree disagreed in places with the traditional tree based on appearance, and it documented extensive gene flow between species. In other words, species in this group exchange genes more than their tidy names suggest.

A point worth getting right: the early Bmp4 and calmodulin experiments were done by changing gene activity in chick embryos. They show that these pathways can alter beak growth in a bird. They do not by themselves prove that these particular changes are what separated each finch species in the wild. The genome-scale work since 2015 has strengthened the case, especially for ALX1, but the full genetic architecture of beak shape is still being worked out.

An analogy helps. Think of the genome as a mixing board with many sliders. Evolution rarely invents new instruments. It nudges existing sliders, such as how much Bmp4 is active, when, and where. Because those sliders are old and shared across birds, a modest change can produce a large visible difference in the beak.

A short timeline of the finch story

Date	Event
September 15, 1835	The Beagle reaches the Galápagos
October 20, 1835	The Beagle sails for Tahiti
1837	Gould identifies Darwin's finch specimens as a distinct group of 12 species
1859	Darwin publishes On the Origin of Species
1866	Mendel publishes his pea experiments
1936	Percy Lowe first applies the name "Darwin's finches"
1947	David Lack's book Darwin's Finches popularizes the name
1977	Drought on Daphne Major; strong selection on beak depth
1981	A male G. conirostris immigrant founds the Big Bird lineage
2002	Grants publish the 30-year study of two Daphne Major species
2004 to 2006	Bmp4 and calmodulin identified in beak development
2015	Genome sequencing of 120 birds finds extensive gene flow and the ALX1 haplotype
2018	Hybrid speciation of the Big Bird lineage published in Science

The naming history deserves a note. Darwin did not call them Darwin's finches. Wikipedia's entry credits Percy Lowe with first applying the term in 1936 and David Lack with popularizing it in his 1947 book.

Why this matters beyond the Galápagos

The finch work changed how biologists think about evolutionary time. Before it, many people treated natural selection as something that had to be inferred from fossils or reconstructed from deep history. After it, selection became something you could estimate in a single bad year.

That has practical consequences. Bacteria evolve antibiotic resistance by the same logic: variation exists, a drug removes susceptible cells, and resistant survivors reproduce. See How Antibiotic Resistance Spreads for the mechanism. Laboratory evolution gives a complementary view, as in The Lenski Long-Term Evolution Experiment, where bacterial populations were tracked for tens of thousands of generations.

Another classic field case is the peppered moth, whose color frequencies shifted with industrial pollution. See Kettlewell's Peppered Moth Experiment Explained. The finches add something the moths do not: a direct link between a measured trait, an environmental change, a death toll, and the heritable shift in the next generation.

Conservation is a second area. If a population can shift its traits within a few generations under strong selection, some species may adapt to rapid change. But the finch data also show that adaptation depends on existing variation. A population that has lost genetic diversity has fewer sliders to move.

Limits of the finch evidence

Strong as it is, the Daphne Major record has boundaries that careful readers should keep in view.

One island, a few species. Daphne Major is small, with a simple ecology. The Grants' own conclusion in 2002 was that long-term studies are needed precisely because rare events matter, and a single site cannot represent every ecosystem.

Microevolution is not the whole story. Measuring a beak shift of half a millimeter shows selection operating within a species. It does not, on its own, explain the origin of major body plans or events such as the Cambrian explosion. Those questions need other kinds of evidence, including fossils and comparative genomics. For one such case, see The Cambrian Explosion: When Animal Life Diversified.

Gene function is partly inferred. As noted above, some developmental experiments used chick embryos rather than finches. They establish what a pathway can do, not the whole history of how a finch species acquired its beak.

Taxonomy is unsettled. The 2015 genome study found that the genetic tree conflicts with trait-based classification in places, and a 2019 analysis by Zink and colleagues, titled "Species Limits and Phylogenomic Relationships of Darwin's Finches Remain Unresolved," argued that species limits and relationships are still unsettled. That is why sources give slightly different species counts.

Old stories are sticky. The notion that Darwin derived evolution from finch beaks in 1835 is a simplification. The record shows a poorly labeled collection, an expert's reinterpretation in 1837, and years of further work.

What this means

Darwin's finches are less a single eureka than a long relay. Darwin collected the birds, Gould recognized their kinship, Lack gave them their name, and the Grants measured selection in action. Geneticists then began to find the molecular dials that make a deep beak or a long one.

The most useful lesson is about timescale and uncertainty. Selection can be strong enough to alter a population within a year, yet it does not push in a fixed direction, and the long-term path stays unpredictable. A beak is good or bad only relative to the seeds on offer, and the seeds depend on the rain.

Frequently asked questions

Q: Why are the Galápagos Islands so useful for studying evolution? A: The islands are volcanic and have never been joined to a continent, so their land animals descended from rare arrivals across open ocean. Newcomers found empty ecological niches and few competitors, which favors adaptive radiation. The islands also differ in climate and vegetation, so the same lineage can face different food supplies. Wikipedia's finch summary places the finches' arrival more than a million years ago.

Q: How many species of Darwin's finches are there? A: About 18 are usually recognized, 17 in the Galápagos and one, the Cocos finch (Pinaroloxias inornata), on Cocos Island. The exact number depends on how species are defined. The 2015 genome study found that gene flow and hybridization blur boundaries, and a 2019 analysis argued that species limits remain unresolved.

Q: Did Darwin name them after himself? A: No. Percy Lowe first applied the name "Darwin's finches" in 1936, according to Wikipedia's entry, and David Lack's 1947 book Darwin's Finches popularized it. Darwin paid the finches little attention during the voyage and did not label them by island, which is why he later needed specimens from FitzRoy and other crew members.

Q: What did the Grants measure after the 1977 drought? A: They measured beak depth in medium ground finches before and after the drought. Teaching materials from the Open University report an average of 9.2 mm before, 9.9 mm in breeding survivors, and 9.7 mm in chicks hatched in 1978. The result showed that large-beaked birds survived better and passed the trait to offspring.

Q: Have scientists watched a new lineage form? A: Close to it. A male Geospiza conirostris from Española arrived on Daphne Major in 1981 and bred with a resident medium ground finch. Their descendants, the Big Bird lineage, bred only among themselves from the second generation, and the authors of a 2018 Science paper estimate reproductive isolation arose in about three generations. Its long-term survival is uncertain.

Q: Which genes control finch beak shape? A: Several have been implicated. Bmp4 expression correlates with deep, broad beaks, a calmodulin pathway with long, pointed beaks, an ALX1 haplotype with beak shape across species, and a region containing HMGA2 with beak size. Some of the earliest tests used chick embryos, so they show what the pathways can do rather than the whole evolutionary history.

Q: Does the finch research matter for medicine? A: It illustrates a principle that applies well beyond birds. Whenever a population varies, the variation is heritable, and an environmental pressure removes some variants, the population shifts. Antibiotic resistance in bacteria follows that same logic, which is why resistance management leans on evolutionary reasoning.

Sources

• Grant, P. R., and Grant, B. R. (2002). Unpredictable evolution in a 30-year study of Darwin's finches. Science 296(5568): 707-711.
• Lamichhaney, S., et al. (2018). Rapid hybrid speciation in Darwin's finches. Science 359(6372): 224-228.
• Lamichhaney, S., et al. (2015). Evolution of Darwin's finches and their beaks revealed by genome sequencing. Nature 518: 371-375.
• Open University. Natural selection on Daphne Major (teaching module summarizing the 1977 drought data).
• Wikipedia. Darwin's finches and Second voyage of HMS Beagle (used for dates, naming history, and the reference list to Abzhanov et al. 2004 and 2006 and Lamichhaney et al. 2016).

---

Read next:

• Charles Darwin: The Idea That Took 20 Years to Publish
• Evolution and Natural Selection — How Life Changes
• The Lenski Long-Term Evolution Experiment