Speciation

Why species split

Populations do not always stay mixed forever. Mountains rise, rivers change course, islands form, climates shift, and groups spread into new habitats. When populations stop exchanging genes, they begin to change in different ways.

Mutation gives each group new variants. Natural selection can favour different traits in different places. Genetic drift pushes small groups apart by chance. None of these forces has to aim at making a new species. The split is a side effect of groups changing separately.

Over time, the differences can build up until the groups no longer interbreed successfully, or do so only rarely. At that point, scientists recognise them as separate species.

What counts as a species

The most widely used definition is the biological species concept. The biologist Ernst Mayr popularised it in 1942. A species is a group of populations whose members can interbreed and produce fertile offspring, and which do not do so with other groups.

It is a useful rule because it focuses on what matters for evolution. If two groups cannot share genes, each one changes on its own from then on.

The rule also has real limits:

• Bacteria and other organisms that do not reproduce sexually. There is no interbreeding to test.
• Fossils. You cannot check whether two extinct animals could have bred, so palaeontologists use body shape instead.
• Groups that hybridise often. Many plants, and some animals such as wolves and coyotes, interbreed where they meet but still stay mostly distinct.

So biologists use several species definitions, depending on what they study. The fuzzy edges are not a flaw in evolution. They are what you would expect if species form gradually. Some pairs of populations are caught partway through the process right now.

How reproductive barriers work

The heart of speciation is reproductive isolation. That means two groups stop exchanging genes, even if they live close together. Biologists sort the barriers into two kinds.

Prezygotic barriers stop mating or fertilisation from happening at all. Postzygotic barriers act after fertilisation. A hybrid forms, but it dies early or cannot breed.

The mule is the classic case. A horse has 64 chromosomes and a donkey has 62. A mule gets half from each parent, so it ends up with 63. When a mule tries to make eggs or sperm, its chromosomes cannot pair up evenly. Mules are almost always sterile. Horses and donkeys can produce offspring, but their genes cannot flow onward, so they remain separate species.

Key idea: A new species forms when gene flow between two groups stops. A sea, a change of food plant, a mismatch in courtship, or a change in chromosome number can each be enough to start the split.

Separated by geography

The most common route starts with physical separation. Biologists call it allopatric speciation, which means “in other places.”

1. One population is divided into two groups.
2. The groups gain different mutations, face different selection, and drift apart by chance.
3. They become more different over time.
4. Breeding between them becomes difficult or stops.

The Isthmus of Panama is a natural experiment on a grand scale. For millions of years, a seaway connected the Caribbean Sea and the Pacific Ocean. As the land bridge rose, it cut the marine life in two. Most estimates put the final closure at roughly 3 million years ago, though geologists still debate exactly when the last connection shut.

In 1993, Nancy Knowlton and colleagues at the Smithsonian Tropical Research Institute studied snapping shrimp of the genus Alpheus on both sides. They found seven pairs of species. For each Caribbean species, the closest relative lived on the Pacific side, exactly the pattern you would expect if one population had been split by the land.

They then put shrimp together in mating trials. Males and females from the same ocean usually accepted each other. Pairs from opposite oceans were far less likely to do so. A few million years apart had been enough to build strong reproductive isolation.

There was one more twist. The pairs were not all equally different. Some had diverged much further than others. The likely explanation is that the seaway closed gradually, so populations living in deeper water were cut off earlier than those in shallow water. Even a single barrier can split species at different times.

Splitting without a mountain in the way

Speciation can also begin while two groups still share the same area. This is called sympatric speciation, and it was once thought to be rare. Two well studied cases show how it can work.

The apple maggot fly

Rhagoletis pomonella is a small fly in North America. Its native food plant is hawthorn. European settlers brought apple trees, and in 1864, in New York’s Hudson Valley, the flies were first recorded laying eggs in apples.

That created two groups living in the same orchards and woods. Apples ripen several weeks earlier than hawthorn fruit, so the apple flies now emerge earlier in the year. The flies mate on or near the fruit they grew up in, and each group prefers the smell of its own plant. Different timing plus different preferences means the two groups rarely meet to breed.

They are not yet full species. Only a few percent of matings cross between them. They are what biologists call host races, a lineage partway through a split. A 2020 study by Hood and colleagues found that host shifts like this one can build reproductive isolation quickly and repeatedly.

Cichlid fish in Lake Victoria

Lake Victoria in East Africa holds more than 500 species of cichlid fish found nowhere else. The lake dried out and refilled, and most of this diversity appears to have arisen in roughly the last 15,000 years.

Colour plays a big part. In many species pairs, females prefer males of a particular colour, such as red or blue. That preference alone keeps species apart even when they share the same rocky shore.

The system also shows that speciation can run backwards. In 1997, Ole Seehausen and colleagues reported that pollution was making the lake water murkier. In cloudy water, females could no longer see male colours clearly, and some species began to hybridise. When the barrier weakens, the species can blend back together.

Instant species in plants

Most speciation is slow. Plants have a fast lane.

Sometimes a hybrid between two plant species doubles its full set of chromosomes. This is called polyploidy. The doubled set can pair up with itself, so the plant is fertile. But it cannot breed successfully with either parent. Reproductive isolation appears in one or two generations.

The best documented case is in the genus Tragopogon, yellow-flowered plants in the daisy family. Three European species were introduced to the Palouse region of eastern Washington and northern Idaho in the early 1900s. They hybridised, and two new polyploid species appeared: Tragopogon miscellus and Tragopogon mirus. The botanist Marion Ownbey described them in 1950. Genetic studies since then show that each has formed more than once, independently, in different towns.

This is not a rare quirk. In 2009, Troy Wood and colleagues estimated that about 15 percent of speciation events in flowering plants and about 31 percent in ferns involve an increase in chromosome sets. Many crops are polyploids too, including bread wheat.

Animals rarely do this, but hybrids can still start new lineages. In 1981, a single male large cactus finch from another island arrived on Daphne Major in the Galapagos, where Peter and Rosemary Grant were studying finches. He bred with a local medium ground finch. Within a few generations, his descendants were breeding only with each other, kept apart by their unusual song and beak size. The Grants and their colleagues described this lineage in 2018. Whether it lasts as a separate species is a question for the future.

Ring species: a split you can walk around

A ring species is a chain of populations wrapped around a barrier. Each population can breed with its neighbours. But where the two ends of the chain meet, they behave like separate species.

The greenish warbler is the best known case. Its populations circle the Tibetan Plateau. Moving around the ring, songs and features change gradually. In central Siberia, where the two ends overlap, the birds have different songs and mostly do not interbreed.

Ring species are useful because they lay out in space what normally happens over time. But real cases are messier than the textbook picture. A 2014 genomic study of the greenish warbler, led by Miguel Alcaide, found several breaks in gene flow around the ring in the past. The chain has not always been continuous. It is a strong example of gradual divergence, but not a perfect one.

How long speciation takes

There is no single answer. Here are the cases from this page, roughly from fastest to slowest:

• One or two generations: polyploid plants like Tragopogon.
• A few generations, still unconfirmed: the Daphne Major finch lineage.
• About 160 years and still partial: the apple maggot fly host races.
• About 15,000 years: more than 500 Lake Victoria cichlids.
• About 3 million years: the Panama snapping shrimp.

For many animals, DNA comparisons suggest splits commonly take hundreds of thousands to millions of years. There is also no single generation when a lineage “becomes” a species. The line is drawn later, by biologists looking back.

A branching tree, not a ladder

Speciation is why the history of life looks like a tree. One line splits into two. Those two lines can split again. Over deep time, many branches appear.

This is also why living species are cousins, not stages in a single chain. Humans are not “more evolved” than birds or fish. We are all current branches of the same tree.

Fossil evidence can show changing forms over time. Genetic evidence can show how closely populations are related. Together they let scientists rebuild the branching.

A simple analogy helps. If a group of people speaking one language splits into separate communities for a long time, each community’s speech drifts. New words appear and pronunciations change. Eventually they no longer understand each other. Speciation is not exactly like language change, but the logic is similar: separation plus time can turn one line into several.

A few things speciation does not mean:

• It does not mean every new species appears suddenly. Most splits are slow.
• It does not mean there is one exact generation when a line “becomes” a species.
• It does not mean one animal gives birth to a different species. Populations split, not individuals.
• It does not mean evolution moves in a straight line from simple to advanced.

Where to go next

To see how small changes add up to splits like these, read micro vs macroevolution. To see more change measured in real time, read observed evolution. For the source of the variation that drives every split, read mutation.