What is Dr. Zubair Khalid's research focus?

Dr. Zubair Khalid specializes in molecular virology, mRNA vaccine development, and computational biology, with a focus on avian pathogens like IBDV and Avian Reovirus.

Where is Dr. Zubair Khalid currently working?

Dr. Zubair Khalid is a Postdoctoral Research Associate at the University of Maryland (UMD), specifically within the Department of Animal and Avian Sciences.

Disruptive Selection: Definition and Examples in Evolution

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-09-20

Disruptive selection is a mode of natural selection in which individuals at both extremes of a trait distribution survive and reproduce better than individuals with intermediate values. Because the middle of the distribution is penalized, the population can split into two or more distinct phenotypic clusters, and this split is one route by which new species form.

Disruptive selection matters because it runs against the intuition that selection always pushes a population toward one "best" form. In many environments the best form is not singular. When resources come in two distinct types, or when hybrids between two incipient forms are poorly adapted, the extremes can win and the average can lose. Understanding this mode explains why some populations stay variable, why bimodal trait distributions appear in nature, and how sympatric speciation (speciation without geographic separation) can begin.

What Disruptive Selection Means

Natural selection acts on heritable variation in traits, and its effect depends on how fitness maps onto trait values. Biologists describe three main modes by their effect on the mean and the variance of a trait.

Directional selection favors one extreme. The mean shifts in that direction, and variance usually shrinks as the favored phenotype becomes common.

Stabilizing selection favors intermediate values. The mean stays roughly the same, and variance shrinks because extremes are removed.

Disruptive selection favors both extremes and penalizes intermediates. The mean may stay near the middle, but variance increases and the distribution becomes bimodal or multimodal, with two peaks separated by a valley of low fitness.

The key diagnostic is the shape of the fitness function, the curve that relates trait value to survival or reproduction. Under disruptive selection the fitness function is U-shaped or has two peaks, so fitness is high at low trait values, dips in the middle, and rises again at high trait values.

Why the Mean Can Stay Put While Variance Grows

A common point of confusion is that disruptive selection does not necessarily move the population mean. If the two extremes are equally favored, gains at one end are balanced by gains at the other, so the average value barely changes. What changes is the spread. The population becomes more variable and increasingly divided into two groups. This is why disruptive selection is sometimes described as selection that reduces the fitness of the average while preserving or increasing overall diversity.

Disruptive Selection Compared With Other Modes

The table below summarizes how each mode affects the mean, the variance, and the shape of the phenotype distribution. It is the fastest way to keep the three modes straight.

Mode of selection	Favored phenotype	Effect on mean	Effect on variance	Effect on distribution shape
Directional	One extreme	Shifts toward that extreme	Usually decreases	Single peak moves
Stabilizing	Intermediate	Stays near current value	Decreases	Single peak narrows
Disruptive	Both extremes	Roughly unchanged if extremes are balanced	Increases	Becomes bimodal or multimodal

A useful memory hook: directional selection moves the peak, stabilizing selection sharpens the peak, and disruptive selection splits the peak.

The Mechanism Behind Disruptive Selection

Disruptive selection arises from two broad biological situations. Both produce the same signature, a fitness valley in the middle of the trait range.

Bimodal Resource Distribution

When a population exploits two distinct resource types, intermediates may be mediocre at both. A bird with a medium beak might crack small seeds less efficiently than a small-beaked bird and large seeds less efficiently than a large-beaked bird. The small and large morphs each specialize and each do well, while the generalist in the middle is outcompeted on both fronts. This is the classic ecological driver of disruptive selection.

Heterozygote Disadvantage

When a trait is controlled by a locus with two alleles, heterozygotes can have lower fitness than either homozygote. This pattern, called heterozygote disadvantage or underdominance, produces disruptive selection on the underlying genotype because the two homozygous classes are favored over the mixed class. The result is a stable or unstable polymorphism depending on the strength of selection, and it can push a population toward fixation of one allele in one subpopulation and the other allele in another.

Both mechanisms are often at work together. A bimodal resource base can favor different genotypes, and heterozygote disadvantage can reinforce the split.

Step by Step: How a Split Begins

A population contains heritable variation in a trait, such as beak depth, body size, or egg color.
The environment offers two distinct ways to make a living, or hybrids between incipient forms perform poorly.
Individuals near the middle of the trait range have the lowest relative fitness.
Individuals near each extreme have higher relative fitness and leave more offspring.
Over generations the middle of the distribution thins out and two peaks emerge.
If the two groups also mate assortatively (preferentially with their own type), gene flow between them falls.
Reduced gene flow allows the two groups to diverge further, which is the early stage of sympatric speciation.

The Classic Example: Black-Bellied Seedcracker and Darwin's Finches

Textbooks often introduce disruptive selection with the black-bellied seedcracker (Pyrenestes ostrinus), an African finch that shows two distinct bill sizes in the same population. Small-billed birds handle soft seeds efficiently, large-billed birds crack hard seeds that small bills cannot, and birds with intermediate bills are less efficient at both tasks. The bimodal bill distribution is a textbook illustration of disruptive selection driven by a bimodal food supply.

A closely studied parallel comes from Darwin's finches in the Galápagos. The medium ground finch (Geospiza fortis) at El Garrapatero on Santa Cruz Island shows a bimodal beak size distribution, with small and large beak morphs. Researchers examined patterns of selection in this population by relating individual beak sizes to interannual recaptures during a prolonged drought. They found strong disruptive selection between the two beak size modes, meaning birds near the middle of the beak size range survived less well than birds at either end [1]. The same study found some evidence of selection against the most extreme beak sizes, possibly from competition with other finch species or gaps in the underlying resource distribution, so selection can simultaneously maintain the current bimodality while constraining how far the two modes drift apart [1].

A later analysis built a series of annual fitness functions from a nine-year mark-recapture dataset of more than 600 medium ground finches in the same bimodal population. This work confirmed that disruptive selection between small and large beak morphotypes was present throughout the study period, not just in the two years originally reported, and that its intensity varied with environmental harshness. Disruptive selection was strongest when precipitation was high during the dry season of the previous year [2].

Why the Finch Case Is So Instructive

The finch system shows disruptive selection acting on a real, measurable trait in a wild population over many years. It also shows that the strength of selection is not constant. Selection fluctuates with climate, which means the balance between maintaining two morphs and merging them can shift from season to season [2]. This temporal variation is a central reason the population remains a dynamic, ongoing case rather than a finished split.

Disruptive Selection and Assortative Mating

Disruptive selection alone can create two phenotypic peaks, but it does not automatically create two species. For divergence to proceed, the two groups must also stop exchanging genes. Assortative mating, the tendency to mate with individuals of the same type, is the bridge.

In the El Garrapatero finch population, the two beak size morphs produce acoustically distinctive songs and mate assortatively by morph. A playback study presented territorial males with songs of their own morph, the other morph, and males from a different locality. Males responded more strongly to same-morph playbacks, with significantly shorter latencies to flight, higher flight rates, and closer approaches to the speaker [3]. This supports the idea that song can act as a behavioral cue for assortative mating, which in turn supports sympatric evolutionary divergence.

Put the pieces together and the logic of disruptive evolution becomes clear. Disruptive selection creates and maintains two morphs. Assortative mating reduces gene flow between them. Reduced gene flow lets the two morphs accumulate further differences. This is why disruptive selection is implicated in sympatric speciation, the formation of new species without a geographic barrier.

How Disruptive Selection Is Detected in Practice

Researchers do not simply look at a bimodal histogram and declare disruptive selection. A bimodal distribution can arise from many causes, including two separate species mixed in one sample, age structure, or sampling artifacts. Detecting disruptive selection requires linking trait values to fitness.

The standard approach is the fitness function. Investigators measure a trait in marked individuals, track survival or reproduction over time, and fit a curve relating trait value to fitness. A U-shaped or two-peaked curve is the signature of disruptive selection. In the finch work, apparent survival was related to beak size across a nine-year dataset to build annual fitness functions [2].

Several practical checks strengthen the inference:

Confirm the trait is heritable, so selection can produce an evolutionary response.
Confirm the two peaks correspond to real ecological differences, such as diet or resource use.
Check whether the population is a single interbreeding group or a mix of cryptic species.
Track selection over multiple years, because a single season can misrepresent the long-term regime [2].

The multi-year finch dataset is a good model here. A single fitness function captures only one selection regime over one time period, so longer-term dynamics require repeated measurement [2]. The finding that disruptive selection persisted across nine years, with intensity tied to rainfall, is far stronger evidence than any single-year result.

Disruptive Selection in Other Systems

Disruptive selection is not limited to finches. The same logic appears wherever two distinct optima exist.

Brood Parasitism and Egg Color

Interactions between brood parasitic common cuckoos and their hosts create strong links between egg phenotype and fitness. Using avian color space models to analyze egg color variation, researchers found pronounced opportunity for disruptive selection on brambling (Fringilla montifringilla) egg coloration. The corresponding cuckoo host race has evolved egg colors that maximize fitness in both sympatric and allopatric brambling populations. The chaffinch (F. coelebs) shows a more bimodal egg color distribution consistent with the evolutionary direction predicted for the brambling [4]. This is disruptive selection driven by an antagonistic interaction rather than by a simple food gradient.

Divergent Selection Across Lineages

Disruptive selection acts within populations, but its logic extends to divergence between lineages. A mitogenome study of flying squirrels and tree squirrels found higher ratios of nonsynonymous to synonymous substitution rates in several mitochondrial genes in the gliding lineage compared with the climbing lineage, suggesting that divergent natural selection acted on the two groups after they split [5]. This is a between-lineage pattern rather than within-population disruptive selection, but it illustrates how selection on different extremes of a functional trait can drive lineages apart.

Balancing Selection and Diversity

Disruptive selection is one of several processes that maintain genetic diversity. Balancing selection maintains variation through mechanisms such as negative frequency-dependent selection, in which a phenotype's fitness rises as it becomes rarer. Distinguishing the genomic signature of negative frequency-dependent selection from other balancing modes such as overdominance remains a significant challenge, and researchers are developing deep learning approaches to tell these signals apart from neutrality [6]. The practical point for students is that disruptive selection is part of a broader family of processes that keep populations variable rather than uniform.

Disruptive Evolution and Polymorphism

A polymorphism is the stable coexistence of two or more distinct forms in a population. Disruptive selection can maintain polymorphism when the two extremes each have an advantage that depends on the environment or on frequency. Because intermediates are penalized, the two forms persist rather than one replacing the other.

This has an important consequence. A population under disruptive selection is not "on its way" to losing diversity. It can sit in a stable two-peak state for a long time, especially if the two resources remain available and assortative mating is imperfect. The finch population at El Garrapatero illustrates this balance. Selection maintains bimodality while also constraining further divergence, producing a dynamic tug of war among selection, assortative mating, and other factors that may alternately promote or limit divergence [1].

When Human Activity Erases Bimodality

The same finch system shows how fragile this balance can be. One study compared the relatively undisturbed El Garrapatero population with the severely disturbed Academy Bay population near the town of Puerto Ayora. El Garrapatero currently shows beak size bimodality tied to assortative mating and disruptive selection, while Academy Bay was historically bimodal but lost that property as local human population density rose. Associations between morphology, bite force, and diet were generally weaker at Academy Bay, possibly because novel foods are used regardless of individual morphology and performance [7]. When the resource base stops being bimodal, the selection that maintained two morphs weakens, and the population can collapse back toward a single mode.

Common Mistakes and Limitations

Students and readers new to the topic make a predictable set of errors. Each has a clean correction.

Confusing disruptive selection with directional selection. Directional selection favors one extreme and moves the mean. Disruptive selection favors both extremes and leaves the mean roughly in place while increasing variance. If the mean shifted and the distribution stayed single-peaked, that is directional selection.

Assuming a bimodal histogram proves disruptive selection. Bimodality can come from sampling two species, from age classes, or from measurement artifacts. Disruptive selection requires evidence that intermediates have lower fitness, which means you need fitness data, not just a distribution.

Thinking disruptive selection always causes speciation. It can initiate divergence, but speciation also requires reduced gene flow, often through assortative mating. Disruptive selection without assortative mating may simply maintain a polymorphism [1][3].

Expecting the two peaks to keep moving apart indefinitely. Selection can also act against the most extreme phenotypes, which constrains divergence. The finch data show exactly this, with disruptive selection between modes coexisting with selection against the largest and smallest beaks [1].

Treating selection strength as constant. Selection intensity varies with climate and other conditions. A single season or year can give a misleading picture of the long-term regime [2].

Overlooking the underlying genetics. Disruptive selection on a phenotype can reflect heterozygote disadvantage, bimodal resource use, or both. The mechanism matters for predicting whether polymorphism will be stable.

Forgetting that individual cases vary. Real populations are influenced by competition, mate choice, climate, and human disturbance at the same time, so no single factor tells the whole story [7].

Quick Review

Disruptive selection favors both extreme phenotypes and penalizes intermediates.
Its signature is a U-shaped or two-peaked fitness function and a bimodal trait distribution.
The mean often stays near the middle while variance increases.
Two main drivers are a bimodal resource distribution and heterozygote disadvantage.
The medium ground finch at El Garrapatero shows strong disruptive selection on beak size, confirmed across a nine-year dataset [1][2].
Assortative mating, often cued by song in finches, is the bridge from disruptive selection to sympatric speciation [3].
Disruptive selection maintains polymorphism and can be weakened when human activity removes the bimodal resource base [7].

Frequently Asked Questions

What is disruptive selection in simple terms?

Disruptive selection is natural selection that favors individuals at both ends of a trait range and works against those in the middle. The result is a population that splits into two distinct groups rather than converging on one average form.

How is disruptive selection different from stabilizing selection?

Stabilizing selection favors intermediate phenotypes and narrows the distribution around the mean. Disruptive selection favors the extremes and widens the distribution into two peaks. They have opposite effects on variance.

What is the classic example of disruptive selection?

The black-bellied seedcracker, an African finch with two bill sizes specialized for soft and hard seeds, is the textbook example. The medium ground finch in the Galápagos is the best-documented wild case, with strong disruptive selection on beak size confirmed over nine years [1][2].

Can disruptive selection cause new species to form?

It can contribute to speciation when it is paired with assortative mating, which reduces gene flow between the two morphs. Disruptive selection alone maintains variation, while reduced gene flow allows the two groups to diverge further [1][3].

What causes disruptive selection?

Two main causes are a bimodal resource distribution, where intermediates are mediocre at both available resources, and heterozygote disadvantage, where the mixed genotype has lower fitness than either homozygous form. Both create a fitness valley in the middle of the trait range.

Does disruptive selection change the population mean?

Usually not much. If the two extremes are favored equally, gains at one end balance gains at the other, so the mean stays near the middle while the variance increases. The distribution changes shape more than it changes position.