Biology › Evolution and speciation › Speciation: how one gene pool becomes two, and why the line is genuinely fuzzy

Speciation: how one gene pool becomes two, and why the line is genuinely fuzzy

A species is not a thing you can point at so much as a statement about who breeds with whom. Once gene flow between two parts of a population stops, mutation, selection and drift work on each part separately, and given enough time the two can no longer interbreed even when they meet. That is speciation, and the interesting arguments are all about where to draw the line.

Before this Natural selection and variation · Evidence for evolution, resistance and drift · Meiosis and the origin of variation

Before you start

Two organisms belong to different species if they look different enough. It is the definition everybody uses in practice and it fails constantly. Male and female eclectus parrots look so unlike each other that they were described as separate species for over a century, and they are the same species. Meanwhile there are pairs of frogs and pairs of gulls that are almost impossible to tell apart on sight and never interbreed. Appearance is evidence about species boundaries; it is not the boundary. What defines the boundary is whether the two gene pools are still connected.

What you should be able to do

State the biological species concept and identify situations in which it cannot be applied.
Explain what reproductive isolation means and why it is the event that creates a species.
Describe allopatric speciation as a sequence, from one gene pool to two.
Describe sympatric speciation and explain why it is harder, with polyploidy as the clearest route.
Classify isolating mechanisms as prezygotic or postzygotic and give an example of each.
Discuss the timescale of speciation and why the boundary between two species is often not sharp.

What a species is, and the cases the definition cannot handle

The definition to quote is the biological species concept: a species is a group of organisms with similar characteristics that can interbreed to produce fertile offspring, and which are reproductively isolated from other such groups. Every word earns its place. 'Interbreed' rules out organisms that merely resemble each other. 'Fertile' excludes the mule, a healthy animal produced by a horse and a donkey and almost always unable to breed. And 'reproductively isolated' does the real work, because a species is defined by the boundary around it rather than by anything inside.

Species: A group of organisms with similar characteristics that can interbreed to produce fertile offspring, and which are reproductively isolated from other such groups.
Gene flow: The movement of alleles between populations, by migration of individuals or of gametes, which keeps their gene pools similar.
Reproductive isolation: The state in which gene flow between two populations has stopped, because of some feature of the organisms or of where and when they breed.
Speciation: The formation of a new species from an existing one, through reproductive isolation followed by genetic divergence of the isolated populations.

The definition is the one to write in an exam, and it is also one you should be able to criticise, because several boards ask for exactly that.

None of these four cases makes the concept useless. What they show is that the interbreeding test is a criterion rather than a measurement, and that biologists in these situations reach for something else — morphology, or DNA sequence similarity — and then argue about the result.

Take them in turn. Asexual organisms — bacteria, many fungi, dandelions reproducing apomictically — never interbreed at all, so prokaryotes in particular are grouped by sequence similarity instead. Fossils cannot be bred, so palaeontologists work from morphology and stratigraphy and treat the resulting species as working hypotheses. Fertile hybrids across a recognised boundary are commoner than the definition suggests: polar and brown bears produce fertile offspring and have exchanged genes repeatedly, and among flowering plants hybridisation between named species is routine. And ring species give the problem in its purest form — a chain of populations in which each interbreeds with its neighbours while the two ends, living side by side, do not. The greenish warbler around the Tibetan plateau is the best-supported case; the herring gull complex in older textbooks no longer holds up to genetic scrutiny as originally described.

So quote the concept, use it, and know that it draws a clean boundary in a world where the boundary is often a gradient. That is what you should expect if species arise gradually, because a process taking thousands of generations must spend most of that time in states that are neither clearly one species nor clearly two.

Allopatric speciation: the barrier does the first half of the work

Speciation needs two things in sequence. First, gene flow between two parts of a population has to stop or fall sharply. Second, the two isolated gene pools have to diverge, until the differences are enough to keep them separate even if the original obstacle goes. Almost every dispute in this topic is about the first step.

In allopatric speciation — from the Greek for 'other homeland' — the first step is geography. A river changes course, sea level cuts an island from the mainland, a mountain range is uplifted, a few individuals are blown a thousand kilometres out to sea. The populations on either side are separated by something neither can cross, and no alleles pass between them.

Frame three is where the biology happens and frame five is where the definition is met. Everything between the barrier appearing and the two forms failing to interbreed is ordinary mutation, selection and drift, running independently in two places at once.

Once separated, the two populations diverge for three reasons at the same time, and a good answer names all three. Different mutations occur in each, because mutation is a chance event and the two populations are no longer sharing them. Different selection pressures act on each, because the two environments are not identical — one side wetter, one side with a different predator, one side with a shorter growing season. And genetic drift acts separately in each, which matters a great deal if one of the populations is small, as it is when an island is colonised by a handful of founders.

Over generations the two gene pools move apart. Because reproductive behaviour, gamete recognition, chromosome structure and development are all under genetic control, changes in those systems accumulate as a by-product. Nothing is selected for incompatibility with the other population — the other population is not there, so it cannot be a selection pressure — but incompatibility arises anyway, as it does for two people editing separate copies of one document.

The test comes when the barrier goes. If the two forms meet again and still interbreed freely, the gene pools merge and there is one species. If they meet and cannot produce fertile offspring, they are two, and the geography no longer matters because the barrier is now inside the organisms. And if they meet and interbreed with reduced success, individuals that mate with the wrong form leave fewer surviving offspring, so anything helping them tell each other apart is favoured — reinforcement, which can complete a separation that isolation started.

Darwin's finches are the textbook case for a reason: a founding population reached the Galápagos from South America, spread to islands separated by water the birds rarely cross, and diverged into species with beaks suited to different foods, some of which now coexist on one island without interbreeding. Snapping shrimps give a cleaner version, because the Isthmus of Panama closed at a known time and there are now pairs of close relatives on either side that will not breed when brought together.

Sympatric speciation: the hard case, and the plant shortcut

Sympatric speciation happens without any geographical separation: the two diverging populations live in the same place throughout. It is harder to get, and for a long time many biologists doubted it happened at all, for an obvious reason — if the two groups are in the same place, gene flow between them can continue, and gene flow is exactly what erases divergence. Something has to reduce it from the start.

In animals the usual candidate is a shift in what the population uses or when it breeds. Rhagoletis pomonella, the apple maggot fly, fed on native hawthorn in North America until apples were introduced in the nineteenth century, and some flies began to use apples instead. The flies mate on the fruit they emerged from, and apples ripen earlier, so the two host races now largely breed at different times and places within one orchard. They are measurably genetically different and interbreed at a low rate: speciation caught partway through, roughly 150 years in. Cichlids in African crater lakes give further cases, where populations using different depths or diets have diverged with assortative mating by colour.

In plants there is a shortcut, and it is the one route by which a species appears in a single generation. Polyploidy is an error in meiosis or mitosis leaving an individual with extra whole sets of chromosomes. An autopolyploid arises within one species: a diploid plant makes a diploid gamete, giving tetraploid offspring. Crossed back to a normal diploid, that tetraploid gives triploids whose chromosomes cannot pair evenly at meiosis, so they are sterile — the tetraploid is reproductively isolated from its own parents the moment it appears, while fertile with other tetraploids.

An allopolyploid is more dramatic still. Two related species hybridise; the hybrid is sterile, because it has one set of chromosomes from each parent and they cannot pair. The chromosome number then doubles, every chromosome acquires a partner, meiosis works again, and the result is a fertile plant that cannot breed with either parent. Spartina anglica, the cord-grass now covering large areas of British salt marsh, arose this way around 1890 from a hybrid between a native and an introduced American species. Bread wheat has three ancestral genomes by the same route, and perhaps a third of flowering plant species originated in a polyploidy event.

	Allopatric	Sympatric
What stops gene flow first	A physical or geographical barrier	Ecological, behavioural or temporal separation, or polyploidy
Where the populations live	Separate areas	The same area
Relative frequency	The common route in animals	Much rarer in animals, common in plants
Speed	Usually thousands of generations	Usually slow, but one generation for a polyploid
Examples	Galápagos finches; snapping shrimps across the Isthmus of Panama; Hawaiian fruit flies	Apple and hawthorn races of Rhagoletis; crater-lake cichlids; Spartina anglica

The isolating mechanisms themselves

Whatever starts the separation, what maintains it is a set of features called isolating mechanisms. They divide cleanly according to whether they act before or after a zygote is formed, and that division is worth learning as a division rather than as a list of nine things.

Any single one of these is enough to keep two gene pools apart. Most real species pairs are separated by several at once, which is why the boundary between well-established species is usually sharp even though it took a long time to build.

Prezygotic mechanisms stop the gametes ever meeting. Geographical isolation separates the populations physically; ecological isolation puts them in different parts of one area, as with Anopheles species breeding in sunlit and in shaded pools; temporal isolation separates them in time, so plants whose flowering peaks are a month apart exchange almost no pollen and corals spawning on different nights do not cross. Behavioural isolation is the commonest in animals, because courtship rituals, songs and pheromones are species-specific and a female that does not recognise the display does not respond. Mechanical isolation means the reproductive structures no longer fit. Gametic isolation means the gametes meet but fail: pollen that will not germinate on the wrong stigma, sperm that cannot bind the surface proteins of the wrong egg.

Postzygotic mechanisms act after fertilisation, and are wasteful for that reason: two gametes have been spent on nothing. Hybrid inviability means the zygote forms but the embryo develops abnormally and dies, because two sets of developmental instructions that evolved apart no longer work together. Hybrid sterility means the hybrid is healthy but cannot make gametes — the mule again, because a horse contributes 32 chromosomes and a donkey 31, so the mule's 63 cannot pair at meiosis. Hybrid breakdown means the first hybrid generation is fine and the second is weak or infertile.

One note on terminology, because the boards word this differently. AQA writes about reproductive separation and expects allopatric and sympatric speciation by name; OCR Biology A asks about the role of isolating mechanisms and expects the prezygotic and postzygotic classification; Cambridge International expects both, treating geographical separation as the allopatric case and ecological and behavioural separation as the sympatric one.

How long it takes, and why nobody can say exactly when it finished

Speciation has no moment. It has a period, usually a long one, during which two populations become progressively less able to exchange genes, and the question 'are they two species yet?' has no better answer than 'partly'.

The timescales vary by orders of magnitude, and the variation is itself informative. An allopolyploid plant is a new species in one generation, because its chromosomal isolation is complete the moment it forms. The apple and hawthorn races of Rhagoletis are perhaps 150 years in and are not there yet. The cichlid species flock of Lake Victoria appears to have radiated within about 15 000 years, which is fast for vertebrates. For most animal groups the usual estimate is between 10⁴ and 10⁶ generations.

What makes the boundary fuzzy is that the mechanisms accumulate gradually and independently. Two populations may already fail to recognise each other's courtship signals while still producing fertile hybrids in captivity; another pair may hybridise readily where their ranges overlap while the hybrids leave few descendants. Grizzly and polar bears interbreed and the cubs are fertile, yet the two have distinct ecologies, morphologies and genomes and have been separate lineages for hundreds of thousands of years. Whether they are one species or two is a question about which criterion you privilege, not about a fact nobody has measured.

Constructing a speciation answer that scores every mark

A species of lizard lives across a large valley. An earthquake diverts a river so that it now runs across the valley floor, and the lizards cannot cross it. Explain how this could lead to the formation of two species. (6 marks)

Start with what the barrier does. The river splits the population, so the two halves are geographically isolated and gene flow between them stops: alleles cannot pass from one gene pool to the other.

Give the three sources of divergence. Different random mutations occur in each population. Conditions on the two sides are not identical, so different selection pressures act and different alleles are advantageous in each. Genetic drift changes allele frequencies independently in each as well.

Say what happens over time. The allele frequencies of the two gene pools become increasingly different over many generations, and the populations become phenotypically and genetically distinct.

Close on the definition. Eventually the differences include features affecting reproduction — courtship behaviour, breeding season, gamete compatibility — so that if the populations meet again they can no longer interbreed to produce fertile offspring. They are reproductively isolated, and are two species.

Two things separate a six-mark answer from a three-mark one: naming all three sources of divergence rather than only selection, and finishing at reproductive isolation. An answer stopping at 'they become very different' has described divergence, not speciation.

TRY IT — Deciding what the data can and cannot show

Two populations of a flowering plant grow on either side of a road. Population X grows on a normal soil; population Y grows on the spoil from an old copper mine and is tolerant of high copper concentrations. Population Y flowers on average three weeks earlier than population X. Hybrids can be produced by hand-pollination and are fertile.

Are X and Y separate species? Justify your answer, and explain what is happening to them.

Check your answer

Not yet, on the biological species concept. Hand-pollination produces fertile hybrids, so the two populations can still interbreed to produce fertile offspring and are not fully reproductively isolated.

But gene flow is already much reduced, by two mechanisms acting together. The three-week difference in flowering time is temporal isolation: when most of Y is shedding pollen, most of X is not receptive. And the two soils impose different selection pressures, so seedlings of the wrong type do badly where they land — copper-intolerant seedlings die on the spoil, and tolerant plants are typically poorer competitors on normal soil.

This is sympatric divergence in progress: adjacent populations, no physical barrier, ecological and temporal isolation. With continued selection the flowering times may separate further, since plants flowering at the same time as the other population waste pollen on crosses that give poorly adapted seedlings.

The examinable move is to answer the question asked — not yet, because fertile hybrids are possible — and then explain why the honest answer is 'not yet' rather than 'no'. Copper tolerance on mine spoil is well studied in Agrostis and Anthoxanthum, and is among the better evidence that sympatric divergence can begin over a few metres.

In the exam

Quote the species definition in full when it is asked for: similar characteristics, interbreed, fertile offspring, reproductively isolated. 'Fertile' and 'reproductively isolated' are separately creditable.
In any speciation answer, say that gene flow between the populations stops. That sentence is the hinge of the whole explanation.
Name all three sources of divergence — different mutations, different selection pressures, genetic drift. Most answers give only selection and lose a mark.
Allopatric means a geographical barrier; sympatric means no geographical barrier. If the question gives you a river or an island, it wants allopatric.
For sympatric speciation in plants, polyploidy is the answer, and the reason it works is that the polyploid cannot pair chromosomes at meiosis with its parent species.
Finish on reproductive isolation. An answer that ends with 'the populations became different' has not described speciation.

Check yourself

Two populations of the same species of bird live on neighbouring islands. Ornithologists find that the two have distinct songs and that females of each population do not respond to the song of the other. When birds from the two islands are kept together in an aviary, a few pairs form and the offspring are healthy and fertile. Are these two species? Explain your answer, and predict what would happen if the two populations came into contact naturally.

Answer

On the biological species concept they are not yet two species, because they can interbreed to produce fertile offspring. The aviary result shows that no postzygotic barrier has developed.

But they are close, and the barrier that exists is behavioural and prezygotic. Song is a species-recognition signal, so if females do not respond to the other population's song, mating between the populations will be very rare in the wild even though it is possible in principle. Reproductive isolation in nature can be effectively complete while the physiological capacity to interbreed remains.

That is what allopatric divergence caught partway looks like. The islands separate the gene pools; different mutations, different selection pressures and drift have made the populations diverge; song is one of the traits that diverged.

On natural contact, one of three things follows. If the song difference is enough, they coexist without interbreeding and are treated as two species. If it is not, hybrids form, gene flow resumes and the gene pools merge back into one. Or hybrids are produced but do less well than either parent type — perhaps their song is intermediate and attracts nobody — so selection favours birds that discriminate most sharply, the song difference sharpens and the separation completes. That third outcome is reinforcement.

The structure that earns marks: apply the definition, name the isolating mechanism and when it acts, then use it to make the prediction.

Questions

Written to the command words the boards use. Try them on paper before opening a scheme: the marks go to points made, not to length.

Question 15 marks

A few individuals of a beetle species are carried by a storm to an island 400 km from the mainland and establish a population there. Describe how this could lead to the formation of a new species.

Mark scheme

B1 the island population is geographically isolated from the mainland one, so gene flow between them stops and alleles no longer pass from one gene pool to the other
B1 different random mutations occur in the two populations, and because they are no longer interbreeding those mutations are not shared
B1 conditions on the island are not identical to those on the mainland, so different selection pressures act and different alleles are advantageous in each population
B1 the island population is small and was founded by a sample of the mainland alleles, so genetic drift changes its allele frequencies by chance as well
B1 over many generations the two gene pools diverge until the differences include features affecting reproduction, so the two forms could no longer interbreed to produce fertile offspring even if they met: they are reproductively isolated and are two species

Question 24 marks

Explain how an allopolyploid plant can become a new species in a single generation.

Mark scheme

B1 two related species hybridise, and the hybrid carries one set of chromosomes from each parent
B1 those chromosomes have no partner to pair with at meiosis, so the hybrid cannot make functional gametes and is sterile
B1 the chromosome number then doubles, so every chromosome acquires a partner, meiosis works again and the plant is fertile with others like itself
B1 crossed back to either parent species it gives offspring whose chromosomes cannot pair evenly, so it is reproductively isolated from both parents from the moment it forms — Spartina anglica arose this way on British salt marsh around 1890

Question 34 marks

Compare prezygotic with postzygotic isolating mechanisms, giving an example of each.

Mark scheme

B1 prezygotic mechanisms act before a zygote forms and stop the gametes ever meeting, whereas postzygotic mechanisms act after fertilisation, on the zygote or on the hybrid
B1 prezygotic examples include temporal isolation, where plants whose flowering peaks are a month apart exchange almost no pollen, and behavioural isolation, where a female does not respond to another population's courtship display
B1 postzygotic examples include hybrid inviability, where the embryo develops abnormally and dies, and hybrid sterility, as in the mule, whose 63 chromosomes cannot pair at meiosis
B1 postzygotic mechanisms are wasteful, because gametes and often an embryo have been spent for nothing, whereas prezygotic mechanisms prevent that expenditure — and most well-established species pairs are kept apart by several mechanisms at once

Question 44 marks

Two forms of a fish live at different depths in one lake. They breed at the same time of year, but each spawns only over the substrate at its own depth, and the two differ in colour. Kept together in a tank, fish of the two forms produce fertile offspring. Suggest whether these are two species, and suggest what is keeping the gene pools apart in the lake.

Mark scheme

B1 on the biological species concept they are not yet two species, because they can interbreed to produce fertile offspring, so no postzygotic barrier has developed
B1 gene flow in the lake is nevertheless much reduced, because the two forms spawn in different places, which is ecological isolation acting before any gametes meet
B1 the difference in colour may act as a behavioural barrier as well, if each form recognises and prefers a mate of its own colour, so matings between the forms are rare even where the fish do meet
B1 this is sympatric divergence in progress: there is no geographical barrier, and if the separation persists the two gene pools may diverge far enough for reproductive isolation to become complete

Question 53 marks

Explain why sympatric speciation is harder to achieve than allopatric speciation.

Mark scheme

B1 in allopatric speciation a physical barrier stops gene flow completely from the start, so the two gene pools change independently of each other
B1 in sympatric speciation the two groups remain in the same place, so individuals can still meet and interbreed and gene flow can continue
B1 gene flow mixes the alleles of the two groups and so erases the divergence as it accumulates, which means an ecological, behavioural or temporal separation — or polyploidy — has to reduce interbreeding from the outset for divergence to build up at all

Question 62 marks

State the biological species concept in full, and state what reproductive isolation means.

Mark scheme

B1 a group of organisms with similar characteristics that can interbreed to produce fertile offspring, and which are reproductively isolated from other such groups
B1 reproductive isolation is the state in which gene flow between two populations has stopped, because of some feature of the organisms themselves or of where and when they breed

Worth remembering

A species: similar characteristics, interbreeds, fertile offspring, reproductively isolated from other groups.
Speciation needs gene flow to stop first, and the isolated gene pools to diverge second.
Divergence has three sources: different mutations, different selection pressures, and genetic drift acting separately in each population.
Allopatric speciation starts with a geographical barrier; sympatric speciation starts with ecological, behavioural or temporal separation, or with polyploidy.
Prezygotic mechanisms stop gametes meeting; postzygotic mechanisms act on the zygote or the hybrid.
Polyploidy is the one route that makes a new species in a single generation, and it is overwhelmingly a plant phenomenon.
The boundary between two species is often genuinely gradual, which is what a gradual process should produce.

← Resistance, drift and the evidence: selection where you can watch it happen · Populations and their limits: what stops the numbers rising →