Sympatric Speciation

Speciation, at its core, is the evolution of reproductive isolation between populations — the buildup of barriers that prevent gene exchange, as introduced in the discussion of Reproductive Isolating Mechanisms. In Allopatric Speciation, a geographic barrier does the first job for free: it stops gene flow mechanically, and isolation accumulates afterward as an incidental byproduct of independent evolution in the two isolated populations. Sympatric speciation removes that free pass. By definition, sympatric speciation is the evolution of reproductive isolation between subpopulations of a single ancestral population that remain in continuous or overlapping geographic contact throughout the divergence process, with no extrinsic barrier to interbreeding. This creates an immediate population-genetic problem. Recombination and mating are powerful homogenizing forces: even weak migration or interbreeding (on the order of one migrant per generation) can, in simple neutral models, prevent populations from differentiating at all. For sympatric speciation to succeed, selection against intermediate or "hybrid" phenotypes must be strong, and — crucially — the genetic architecture must allow disruptive selection to become linked with assortative mating, habitat choice, or timing of reproduction. Without that linkage, selection can maintain a genetic polymorphism (two morphs coexisting) without ever producing two species, because individuals of different morphs still mate randomly and constantly regenerate poorly adapted intermediates. Three broad routes achieve this linkage, and they form the organizing scope of this article. First, disruptive selection on resource use (e.g., seed size, host plant) combined with assortative mating based on the same trait or a correlated one — this is the "magic trait" or pleiotropy route. Second, host or habitat shifts, where a change in where organisms feed, mate, or breed automatically reduces encounter rates between shifted and unshifted individuals — allochronic or microhabitat isolation, a special case of the first route driven by ecology rather than mate choice per se. Third, polyploidy — a sudden doubling (or odd multiplication) of chromosome number — which creates near-instantaneous, complete reproductive isolation from the parental diploid population in a single generation, sidestepping the gradual-divergence problem entirely.

Route 1: Disruptive selection plus assortative mating. Imagine a population exploiting a resource that varies continuously — say seed hardness, or host-plant chemistry. If fitness is highest at two extremes and lowest for intermediate phenotypes (a classic disruptive selection regime), selection alone will maintain variance but will not, by itself, split the population. The population remains one interbreeding gene pool, continually generating intermediates through mating between the two extreme morphs. Speciation requires an additional ingredient: assortative mating, meaning individuals preferentially mate with others of similar phenotype. If the trait under disruptive selection is genetically or developmentally linked to the mating cue itself — a "magic trait" — then selection on resource use automatically increases assortative mating as a side effect, without needing a separate, independently evolving mate-choice gene. Beak size in a seed-eating bird is a textbook example of a magic trait: it determines both what food you can eat and, if mate choice tracks body/beak size, whom you tend to court. Population genetic models (John Maynard Smith, 1966; Dieckmann & Doebeli, 1999; Kondrashov & Kondrashov, 1999) show that this route to sympatric speciation is mathematically possible but demanding. It requires:

• Strong disruptive selection (low fitness of intermediates relative to extremes)
• A genetic correlation, or physical linkage, between the ecological trait and the mating trait
• Enough standing genetic variation for the population to actually reach the two fitness peaks
• Low enough recombination between the loci controlling the ecological trait and those controlling mate choice that the correlation is not immediately broken up each generation

If linkage is weak, recombination each generation re-shuffles ecological and mating alleles, and the correlation needed to keep assortment aligned with adaptation collapses — the population stays polymorphic but never splits. This is why "one-allele" mechanisms (where the same allele change increases assortment in both morphs, as with a magic trait) evolve far more easily under models than "two-allele" mechanisms requiring separate, coordinated evolution of choice alleles in each morph.

Route 2: Host or habitat shift. A subset of a population begins using a new resource — a new host plant, a new microhabitat, a shifted breeding season. Critically, mating in many phytophagous insects occurs on the host: courtship, mate search, and copulation happen on the same plant where larvae feed. This means a host shift automatically creates assortative mating as a mechanical byproduct of habitat choice — individuals that shift hosts encounter and mate mostly with other shifted individuals, without any evolution of an explicit "prefer similar mates" allele. Selection on host-related traits (diapause timing, host-finding cues, digestive physiology) then proceeds somewhat independently in the two host races, while gene flow between them is suppressed passively by reduced encounter rate. Over time, habitat isolation is reinforced by divergence in diapause timing (an allochronic component — the two races literally emerge from pupation at different times of year, tracking their different host's phenology) and eventually by genetic incompatibilities that accumulate as a side effect (a process converging with allopatric mechanisms once initial isolation is strong).

Route 3: Polyploidy. This route is mechanistically distinct and does not require disruptive selection at all. Polyploidy arises when chromosome number is multiplied, most commonly through:

• Autopolyploidy — chromosome doubling within a single species, typically via failure of meiotic reduction (production of unreduced, diploid gametes) followed by fusion of two diploid gametes to form a tetraploid zygote.
• Allopolyploidy — hybridization between two different species followed by chromosome doubling, which restores fertility to an otherwise sterile hybrid by giving every chromosome a pairing partner during meiosis.

The reproductive isolation is immediate and nearly absolute: a tetraploid (4n) individual that mates with a diploid (2n) parent produces a triploid (3n) offspring. Triploids have an odd number of chromosome sets, so during meiosis chromosomes cannot pair up evenly into two viable gamete sets; the result is grossly unbalanced gametes and, almost universally, sterile or highly inviable offspring. This "triploid block" means that as soon as a fertile tetraploid lineage exists, gene flow with the diploid parent is sharply reduced in a single generation — postzygotic isolation appears essentially instantly, without geographic separation, without disruptive selection, and without gradual accumulation of incompatibility alleles.

Whichever route operates, the outcome tested against real populations is the same: can we show that isolation evolved, and is currently maintained, without any geographic separation ever having occurred? That evidentiary bar — ruling out a cryptic past allopatric phase — is exactly what the classic studies below were designed to meet.

Apple maggot flies, Rhagoletis pomonella. This is the most thoroughly documented candidate for sympatric, host-race origin of a new species in progress. Rhagoletis pomonella is native to North America, historically breeding exclusively on hawthorn (Crataegus) fruit. Around the mid-1800s, following the introduction of domesticated apple trees, a subset of the fly population began breeding on apples instead — apple and hawthorn trees often grow interspersed, sometimes in the same orchard hedgerow, so this shift involved no geographic separation at all. Guy Bush proposed in 1969 that this host shift could drive sympatric speciation, and decades of subsequent work by Jeffrey Feder, Bruce McPherson, and colleagues built the empirical case:

• Assortative mating by host. Flies court and mate on the fruit of their natal host. Mark-recapture and direct observation studies show that apple-origin flies overwhelmingly mate on apples and hawthorn-origin flies on hawthorns, producing strong assortative mating as a passive consequence of habitat/host fidelity rather than an evolved mate-preference gene.
• Allochronic isolation from phenology differences. Apples fruit roughly 3–4 weeks earlier than hawthorn in most regions. Because Rhagoletis larvae overwinter as pupae and diapause timing is heritable, selection has shifted the emergence time of apple-race flies earlier to match apple fruiting phenology. This reduces temporal overlap in adult emergence between the two host races, on top of the host-based spatial assortment.
• Genetic differentiation despite ongoing sympatry. Allozyme and later genomic studies (Feder et al., 1988 and onward) found consistent allele-frequency differences between apple-race and hawthorn-race flies at several loci, including genes clustered in chromosomal inversions that resist recombination — exactly the kind of "genomic islands" of differentiation predicted by models requiring linkage between ecological and reproductive-isolation loci. Gene flow between races is reduced (estimated around a few percent per generation) but not zero.

This system is celebrated precisely because it captures speciation-in-action: two host races, genetically differentiated, partially reproductively isolated by habitat and timing, still capable of interbreeding at low rates, in full geographic sympatry. It has not (yet) produced two fully distinct species — which is itself an important lesson: sympatric speciation is a process with intermediate, observable stages, not an instantaneous event.

Crater lake cichlids. Small, geologically young, and closed volcanic crater lakes in Cameroon and Nicaragua provide some of the cleanest natural tests because their basins are physically too small and uniform to have permitted cryptic geographic subdivision. Lake Apoyo and Lake Xiloá in Nicaragua, colonized recently (within roughly the last few thousand to ten thousand years) by a single ancestral Amphilophus species, each now contain multiple sympatric cichlid species pairs differing in body shape, jaw morphology, and diet — some feeding on benthic invertebrates, others on open-water plankton. Phylogenetic and population-genomic analyses (Barluenga et al., 2006, and subsequent genomic work by Axel Meyer's group) show that within-lake species are each other's closest relatives (monophyletic within a lake), rather than each being more closely related to fish in a different lake — the pattern expected if divergence occurred in situ, sympatrically, rather than by repeated independent colonization from outside. Disruptive selection on trophic morphology, combined with assortative mating linked to body shape and habitat depth, is the inferred mechanism, echoing the "magic trait" model.

Polyploid speciation in Tragopogon (goatsbeards). This is the single best-documented case of observed, repeated, ongoing sympatric (in fact instantaneous) speciation via allopolyploidy in a wild plant. Three Eurasian Tragopogon species — T. dubius, T. pratensis, and T. porrifolius — were introduced to the Palouse region of the western United States in the early 1900s and now grow intermixed. Where their ranges overlap, they hybridize, and two new allopolyploid species have formed independently and repeatedly since roughly the 1950s: T. mirus (from T. dubius × T. porrifolius) and T. miscellus (from T. dubius × T. pratensis). Botanist Marion Ownbey first documented these new polyploids in 1950, only decades after the parent species had been introduced — a speciation event essentially observed within living human memory. Molecular work (Soltis & Soltis, and later Doug Soltis's lab, using chloroplast DNA and nuclear markers) confirmed these polyploids arose multiple independent times from different hybridization events, not from one single ancestor, and demonstrated they are reproductively isolated from both parent diploids by the triploid block described above.

Example 1 — a host-race genetics problem. Suppose in a Rhagoletis-like system, diapause timing is controlled largely by a single locus with two alleles: E (early emergence, favored on apple, which fruits early) and L (late emergence, favored on hawthorn, which fruits late), with E dominant to L. A population is initially a single panmictic hawthorn-feeding population, fixed for L/L. A mutation introduces the E allele at low frequency. Suppose apple trees are newly available, and flies with genotype EE or EL now have a fitness advantage of 1.4× on apple relative to hawthorn (because their emergence timing matches apple fruiting), while LL flies retain their ancestral 1.0× fitness advantage on hawthorn. If we cross a heterozygous EL fly (attracted, let's assume, in equal measure to both fruits before host fidelity evolves further) with another EL fly, the offspring genotype ratio by Mendelian segregation is: ``` EE : EL : LL = 1 : 2 : 1 ``` Given E dominance, phenotypically that's 3 early-emerging : 1 late-emerging. If early emergers preferentially encounter and mate on apple (habitat/timing-based assortment), then of the 4 offspring classes, the 3 "early" individuals (EE and EL) funnel disproportionately into the apple-fruiting population while the 1 "late" (LL) funnels into hawthorn. Over successive generations, this genetic sorting — driven by phenology rather than any deliberate mate discrimination — enriches the apple subpopulation for E and depletes it of L, and vice versa on hawthorn, even though flies of both host races remain capable of full interbreeding and share the same gene pool at every unlinked locus. This illustrates precisely why real Rhagoletis studies find differentiation concentrated at particular loci (often in inversions) rather than genome-wide: only loci linked to the ecological/phenological trait diverge, while the rest of the genome continues to be homogenized by residual gene flow.

Example 2 — a polyploidy arithmetic problem. A diploid plant species has 2n = 14 chromosomes (n = 7). Occasionally, due to a meiotic error, an individual produces unreduced (2n = 14) gametes instead of normal reduced (n = 7) gametes. Two such unreduced gametes fuse. ``` 2n (14) + 2n (14) = 4n (28) ``` The resulting tetraploid offspring has 28 chromosomes. Now suppose this tetraploid, growing right next to its diploid parent population, is pollinated by a normal diploid individual (contributing an n = 7 gamete): ``` 4n gamete (14) + n gamete (7) = 3n (21), a triploid ``` During meiosis in this triploid, 21 chromosomes cannot be evenly divided into two balanced sets of homologous pairs — most chromosomes lack a consistent pairing partner, producing wildly unbalanced gamete chromosome numbers and, in the vast majority of cases, inviable or sterile offspring. This is the triploid block. Consequently, gene flow from the diploid parent population into the tetraploid lineage is sharply curtailed at the very first generation. If, however, two tetraploid individuals (4n = 28) mate with each other, their gametes (n = 14 each) combine normally: ``` 14 + 14 = 28 (4n), fully viable and fertile ``` So a self-sustaining, reproductively isolated tetraploid lineage can be founded from a single unreduced-gamete event (or, more robustly in nature, from repeated independent origins, as documented for Tragopogon mirus), with isolation from the diploid parent essentially complete from generation one — no geographic separation, no gradual accumulation of incompatibilities, and no disruptive selection required.

Ernst Mayr, the most influential architect of the mid-twentieth-century "Biological Species Concept," was for decades the most prominent skeptic of sympatric speciation, arguing in his 1963 book Animal Species and Evolution that gene flow would almost always overwhelm local disruptive selection before isolation could evolve, and that essentially all animal speciation was allopatric. This view held enormous influence, in part because it was theoretically conservative and matched most well-studied bird and mammal cases. The counter-tradition was built by researchers who took theory and specific natural systems seriously despite Mayr's skepticism. John Maynard Smith's 1966 population-genetic model showed sympatric speciation was mathematically possible under disruptive selection with the right genetic architecture, giving the idea formal legitimacy. Guy Bush's fieldwork and 1969 hypothesis on Rhagoletis host races gave it an empirical foothold. Decades later, Jeffrey Feder's long-term genetic studies of apple maggot flies, and Ulf Dieckmann and Michael Doebeli's 1999 adaptive-dynamics models (showing sympatric speciation via disruptive selection and assortative mating under broader conditions than previously assumed), shifted the consensus from "essentially impossible" to "possible under specific, identifiable, testable conditions" — precisely the nuanced position taught today.

Agricultural pest management. Understanding Rhagoletis-style host-race formation directly informs how agronomists anticipate new pest races. When a pest insect shifts to a novel crop, the same evolutionary logic predicts rapid divergence in emergence timing and host fidelity, meaning pesticide-resistance genes or life-history changes can evolve and spread within a pest population sympatrically, sometimes faster than expected from models assuming free gene flow.

Crop breeding and polyploidy. Deliberate induction of polyploidy (using colchicine to block spindle formation during meiosis, producing unreduced gametes) is a routine plant-breeding technique. Bread wheat (Triticum aestivum) is itself a natural allohexaploid, formed from two successive interspecific hybridization-and-doubling events combining three ancestral grass genomes; breeders exploit the same mechanism to create new triploid seedless fruit (like commercial seedless watermelon and banana) by deliberately generating sterile triploid hybrids from tetraploid × diploid crosses.

Conservation genetics and cryptic species. Recognizing sympatric or parapatric divergence in progress (partial host races, incipient ecotypes) matters for conservation because a single named species may actually comprise multiple partially isolated lineages with different ecological requirements — management plans that ignore this can inadvertently homogenize genetically and ecologically distinct populations, or fail to protect a lineage on the brink of independent species status. This same logic reappears when biologists study Hybrid Zones, where partially isolated lineages meet and gene flow is measured directly.

Understanding the genomic architecture of speciation. Modern genomic surveys of Rhagoletis, cichlids, and other sympatric or parapatric systems have driven the broader field's focus on identifying "genomic islands of divergence" — chromosomal regions (often near inversions) that resist gene flow while the rest of the genome remains homogenized. This concept is now central to The Genetics of Speciation, extending far beyond sympatric cases to speciation generally.

Ask the tutor: How does the strength of disruptive selection needed for sympatric speciation change as the migration rate between diverging morphs increases?