Ring Species

The paradox of a species that is also two species

Imagine you set out walking from the southern Himalayas and follow a single kind of small green songbird northward. As you go, the birds change a little from valley to valley — the song gets a touch longer, the wing bar a shade brighter — but every bird you meet interbreeds happily with the birds in the next valley. You loop all the way around the Tibetan Plateau, thousands of kilometers, and come back down the other side. When you arrive back near where you started, you find two kinds of these birds living side by side in the same forest — and they completely ignore each other as mates. They are two species. Yet you got from one to the other in an unbroken chain where every link interbred with its neighbor.

That is a ring species, and it is a genuine logical knot. The biological species concept says a species is a group of populations that actually or potentially interbreed. By that rule, population A interbreeds with B, B with C, C with D, all the way around to the population at the closure point — so transitively they should all be one species. But the two ends, sitting in the same place, do not interbreed, so they are two species. Both statements are true at once. The ring is the situation that exposes "species" as a snapshot of a process, not a sharp natural kind.

The reason this matters far beyond birdwatching is that it converts an argument about time into a picture in space. Speciation normally happens over tens of thousands to millions of years and you can never watch the whole thing. A ring species takes that long temporal sequence — variety, divergence, partial isolation, full isolation — and lays it out as a geographic gradient you can survey in a single field season. The continuum is right there to walk along.

How a ring species forms, step by step

The mechanism has four ingredients, and all four must line up:

1. A barrier with habitat wrapped around it. Start with an uninhabitable region — a high plateau, a desert, a deep valley, an ice sheet — that has livable terrain forming a loop around its rim. The Tibetan Plateau (too high and cold for greenish warblers) and California's Central Valley (too hot and dry for the woodland Ensatina) are the canonical barriers.
2. Expansion around two arms from a single source. An ancestral population at one point on the rim expands. Because it cannot cross the barrier, it spreads in two directions — say up the western flank and up the eastern flank. These two fronts are now diverging populations that share an origin but are increasingly separated by distance.
3. Gradual divergence by isolation-by-distance. Along each arm, neighboring populations still swap genes, so change is gradual. But genes do not flow instantly across thousands of kilometers; they diffuse stepwise from neighbor to neighbor. Drift, local adaptation, and (crucially for the warbler) sexual selection on song push the two arms apart in slightly different directions. By the time you are far around the ring, accumulated differences are large even though no single neighbor-to-neighbor step is.
4. Secondary contact and reproductive isolation at closure. The two arms finally meet on the far side of the barrier. If they have diverged enough that they no longer recognize each other as mates — or that hybrids are inviable or infertile — the closure point hosts two species living in sympatry (the same place) without merging. The ring is complete.

The key engine is isolation by distance: gene flow is local, so distant populations are genetically far apart even when the whole chain is connected. The barrier is what forces "distant" to also mean "back in the same place." Without the loop, the extremes of a gradient never get to test whether they have become reproductively isolated.

The greenish warbler — speciation you can hear

The greenish warbler, Phylloscopus trochiloides, is the cleanest documented ring species and the one where you can literally hear the divergence. The ancestral population sits in the southern Himalayas. From there the birds expanded north along two arms — a western arm up through Afghanistan, the Caucasus, and into European Russia (the form viridanus), and an eastern arm up through China and Mongolia into Siberia (the form plumbeitarsus). The two arms close in central Siberia.

Darren Irwin and colleagues, working through the late 1990s and 2000s, recorded songs all the way around the ring. The result is striking: song complexity changes gradually along each arm, and the two arms have pushed song in different directions. Southern birds sing short, simple songs; both northern forms sing longer, more complex songs, but built in different ways — the eastern (plumbeitarsus) form strings together many short units, the western (viridanus) form uses fewer, longer units. Where the two meet in Siberia, females respond to their own form's song and ignore the other's. Playback experiments confirm the birds discriminate. There is essentially no interbreeding at the closure point even though the two forms are linked by an unbroken interbreeding chain around the south.

Genetic work (mitochondrial and later genomic markers) backs up the geographic story: differentiation increases steadily around the ring, consistent with a single southern origin and two diverging arms. The greenish warbler is therefore the rare case where behavior (song), morphology (plumage), and DNA all agree that you are looking at a continuum of speciation frozen in space.

The Ensatina salamanders — a ring around the Central Valley

In North America the star example is the lungless salamander Ensatina eschscholtzii, which forms a horseshoe of subspecies around California's hot, dry Central Valley. From a northern ancestral stock, the salamanders spread south down two arms: a coastal arm through the wetter Coast Ranges (ending in the plain-colored E. e. eschscholtzii) and an inland arm down the Sierra Nevada foothills (ending in the boldly blotched E. e. klauberi). Along the way the inland forms evolved striking warning-like color patterns while the coastal forms stayed plain.

The two terminal forms come back into contact in the mountains of southern California. There they overlap with very little successful interbreeding — hybrids are rare and show reduced fitness — so they behave as separate species, even though they are linked by an interbreeding chain that runs all the way around the valley to the north. Robert Stebbins described the system in the 1940s–1950s, and David Wake's group spent decades on the genetics. Their molecular work added an important honest caveat: the Ensatina ring is not perfectly smooth. There are spots where gene flow was interrupted in the past and zones of secondary contact, so the "ideal" textbook ring is really a ring with some breaks and reconnections — which is exactly what you would expect from a transient, messy evolutionary process.

Ring species vs ordinary speciation patterns

Property	Ring species	Plain cline	Allopatric speciation
Geometry	Chain bent into a loop by a barrier	Linear gradient across a range	Two ranges split by a barrier
Gene flow	Continuous around the ring	Continuous along the gradient	Cut off between the two groups
Terminal overlap?	Yes — the defining feature	No, ends never meet	Only on secondary contact, if it happens
Reproductive isolation	Full at closure, none between neighbors	None (one species throughout)	Builds in isolation, tested at contact
Speciation visible as	A gradient in space	Not speciating	A before/after in time
Stability	Fragile — breaks if a link dies out	Stable, common	Stable once split
How common	Very rare (a handful accepted)	Ubiquitous	The default mode of speciation
Canonical example	Greenish warbler, Ensatina	House sparrow body-size cline	Darwin's finches on separate islands

The scale and tempo behind the loop

• Ring circumference — thousands of kilometers. The greenish warbler ring encloses the Tibetan Plateau and runs on the order of 3,000+ km around its rim. The Central Valley that the Ensatina ring wraps is about 720 km long and 60–100 km wide — small for a bird, enormous for a salamander that may move only meters per generation.
• Divergence time — tens of thousands to a few million years. The greenish warbler complex is estimated to have spread around the plateau over roughly the last few hundred thousand to a couple of million years, with the song and genetic differences accumulating across that window. The Ensatina radiation around the Central Valley is older, on the order of several million years, which is part of why its ring shows more breaks.
• Gene flow is local, not global. Greenish warblers, like most small passerines, typically disperse only a few kilometers to tens of kilometers from their birthplace per generation. That short dispersal is what makes isolation-by-distance work: genes diffuse neighbor-to-neighbor rather than mixing the whole ring each generation.
• Song complexity, not just looks, does the isolating. In the warbler, the northern songs are markedly longer and more syllable-rich than the simple southern ancestral song. Mate choice keying on song means the behavioral isolation can outrun the genetic divergence — birds that share most of their genome still won't pair if the song is wrong.
• Hybrid fitness at closure is low but nonzero. In Ensatina, hybrids between the terminal forms occur occasionally in the southern contact zone but show reduced viability and fertility, which is why the two ends stay distinct rather than blending back together.

Where else ring-like patterns turn up

• Greenish warbler (Phylloscopus trochiloides). The gold standard, ringing the Tibetan Plateau, isolated by song. Western viridanus and eastern plumbeitarsus coexist without interbreeding in central Siberia.
• Ensatina eschscholtzii salamanders. A horseshoe of color-pattern subspecies around California's Central Valley; coastal eschscholtzii and inland klauberi meet in the south with little hybridization. The best-studied case, and the one that taught biologists how messy real rings are.
• Herring gull / lesser black-backed gull complex. The original textbook example, said to ring the Arctic so that British herring gulls and lesser black-backed gulls are the non-interbreeding ends of a circumpolar chain. Modern genetic work (Liebers et al., 2004) showed it is not a clean single ring — it involves multiple colonization events — so it now serves as a cautionary tale rather than a confirmed case.
• Asian greenish warbler relatives and other Phylloscopus. Several warbler complexes around the same plateau show ring-like or near-ring divergence, suggesting the geometry repeatedly nudges these birds toward the pattern.
• Euphonia and Acacia ant ring-like systems, plus the deer mouse (Peromyscus) and certain Drosophila chains have all been proposed at various times; most turn out, on close inspection, to be partial or broken rings — which reinforces how special the geometry has to be.

Common misconceptions

• "A ring species is just a really wide species with a lot of variation." No — the defining feature is the terminal overlap with reproductive isolation. Lots of species vary across their range (that's a cline); only a ring species bends the variation into a loop so the two extremes meet in the same place and fail to interbreed there.
• "The whole ring is connected, so it must be a single species." This is the paradox, not a refutation. Adjacent populations interbreed, but interbreeding is not transitive when divergence accumulates around a loop. At the closure point you genuinely have two species, so the system is one species and two species depending on where you stand.
• "The herring gull is the classic ring species." It was the classic textbook example for decades, but genetic analysis broke it: the circumpolar gulls are a web of separate colonizations, not one clean ring. Use the greenish warbler or Ensatina instead.
• "Ring species prove the biological species concept is wrong." They expose its limits, not its uselessness. The concept works fine for most well-separated species; ring species (and other intermediate cases) just show that "species" has fuzzy edges when you catch evolution mid-act. That fuzziness is a feature of a continuous process, not a bug in the theory.
• "A ring species is permanent." It's one of the most fragile arrangements in nature. If any middle population in the chain goes extinct, the loop breaks into two ordinary disconnected species and the continuum is gone. The pristine, unbroken ring is a brief evolutionary moment, which is precisely why genuine examples are so rare.
• "Speciation needs a hard physical barrier between the two new species." The ring species shows the opposite: the two end species are formed while still connected by gene flow around the long way. Isolation by distance plus divergent selection can do the job without ever fully cutting the chain — until the ends meet and discover they no longer match.