Positive Frequency-Dependent Selection

In the intricate dance of evolution, an organism's success is often relative, determined not by its inherent strengths alone, but by the prevalence of its traits within its community. This concept, known as frequency-dependent selection, challenges us to look beyond fixed fitness values. While being rare can sometimes be an advantage, this article focuses on the opposite and equally powerful scenario: What happens when conformity pays, and being common is the key to survival? We will uncover the profound consequences of this "safety in numbers" principle. This article first explores the core ​​Principles and Mechanisms​​ of positive frequency-dependent selection, distinguishing it from density dependence and contrasting its "winner-takes-all" dynamic with the diversity-preserving nature of negative frequency-dependent selection. We will then journey through its far-reaching implications in the ​​Applications and Interdisciplinary Connections​​ chapter, seeing how this single idea explains everything from the warning colors of butterflies to the universal language of our own genetic code.

In our introduction, we touched upon the idea that in the grand theater of evolution, an organism's success isn't always an absolute, fixed value. Sometimes, its fate is tied to a rather democratic, or perhaps undemocratic, principle: how common or rare its particular strategy is among its peers. This is the essence of ​​frequency-dependent selection​​. But before we dive into its fascinating consequences, let's be clear about what we mean.

Imagine you're a biologist studying a species of animal that comes in two distinct color forms, or morphs. You notice that the reproductive success of each morph seems to change. What could be causing this? Two very different possibilities immediately come to mind.

First, it could be a matter of total population size. As the total number of individuals—the ​​density​​—increases, resources like food and nesting sites become scarcer for everyone. This leads to what biologists call ​​density dependence​​, where the per-capita success rate of all individuals tends to fall as the population gets more crowded. If you were to conduct an experiment where you keep the proportion of the two morphs equal (say, 50-50) but vary the total number of individuals, you would see the reproductive output of both morphs decline as the enclosure gets more packed.

But there's a second, more subtle possibility. What if success depends not on the total size of the crowd, but on who is in the crowd? What if an individual's fitness hangs on the relative frequency of its own morph? This is ​​frequency-dependent selection​​. Here, you would design your experiment differently. You would keep the total density constant but change the proportions of the two morphs. If you see that a morph's fitness changes as its own percentage in the population changes, you have uncovered frequency dependence. It’s a crucial distinction: one is about the size of the party, the other is about your popularity at the party.

Once we've established that frequency matters, we find that it can work in two opposing ways.

The first and perhaps more intuitive version is ​​negative frequency-dependent selection (NFDS)​​. This is the "hipster effect" in nature: an advantage to being rare. Imagine predators that form a "search image" for the most common prey type. They get very good at spotting the common morph, while the rare one tends to slip by unnoticed. The rare morph has higher fitness, so its numbers increase. But as it becomes more common, the predators' search image might shift, and its advantage fades. This process naturally maintains variety, or ​​polymorphism​​, because whichever type becomes rare gets a boost. The system tends to settle into a stable balance where both types coexist. Mathematically, if we plot the fitness of a type against its frequency, we see a downward slope: the more of you there are, the worse you do.

But what about the opposite scenario? What if there's a penalty for being different and an advantage to being common? This is our main subject: ​​positive frequency-dependent selection (PFDS)​​. This is the "safety in numbers" principle, or the "tyranny of the majority." As we're about to see, its consequences are dramatically different.

Let's imagine a classic scenario involving aposematic, or warningly-colored, prey. Picture a species of butterfly that comes in two vibrant, unpalatable varieties: Pattern A and Pattern B. A young, inexperienced predator tries to eat a butterfly with Pattern A, has a disgusting experience, and quickly learns to avoid that pattern in the future.

Now, suppose Pattern A is very common and Pattern B is very rare. Predators in the area have had many opportunities to learn that A is nauseating. An A-butterfly flies around with a high degree of protection. But what about the rare B-butterfly? Since it's rare, few predators have ever encountered it. A predator sees this novel pattern, has no prior bad experience, and is likely to attempt an attack. Being rare is dangerous. The more common your warning signal, the more "educated" the local predator population is, and the safer you are.

This is a perfect example of PFDS. The fitness of a morph increases with its own frequency. We can capture this with a beautifully simple mathematical model. Let's say the frequency of type A is $p$ and the frequency of type B is $1-p$. The fitness of each type might be given by a pair of linear functions:

In these equations, $w_A(p)$ and $w_B(p)$ are the fitnesses of type A and type B, respectively, when the frequency of type A is $p$. The constant $s$ represents the strength of selection. As the frequency $p$ of type A increases, its fitness $w_A(p)$ increases, while the fitness of type B, which has frequency $1-p$, decreases. This direct relationship between a type's frequency and its fitness is the mathematical signature of positive frequency-dependent selection.

So what happens in a population governed by this "winner-takes-all" dynamic? Let's analyze the difference in fitness, $w_A(p) - w_B(p)$. According to our simple model, this is $s(2p - 1)$. Selection will favor type A if this difference is positive, and type B if it's negative.

The outcome hinges entirely on the value of $p$ relative to one special point: $p^* = \frac{1}{2}$.

• If the frequency of type A, $p$, is greater than $\frac{1}{2}$, then $2p-1$ is positive, and type A has higher fitness. Selection will push its frequency even higher, towards $1$.
• If $p$ is less than $\frac{1}{2}$, then $2p-1$ is negative, and type B has higher fitness. Selection will drive the frequency of A down, towards $0$.

The frequency $p^* = \frac{1}{2}$ is an equilibrium, but it's an ​​unstable equilibrium​​. It's like balancing a ball on the very top of a hill. Any slight puff of wind will send it rolling away. In our population, any small random fluctuation in frequency away from $\frac{1}{2}$ will be amplified by selection, sending the population hurtling towards a state where only one type remains.

This leads to two crucial concepts. The first is ​​bistability​​: the system has two stable states—all-A ($p=1$) or all-B ($p=0$)—and the one it ends up in depends entirely on which side of the tipping point it started. This is a profound example of ​​historical contingency​​ in evolution; the initial conditions dictate the final outcome. The second concept is the ​​invasion threshold​​. For a new mutant type to succeed, it's not enough for it to be slightly better in some absolute sense. It must be introduced into the population at a frequency above this critical threshold. Below the threshold, it is doomed by the disadvantage of rarity.

We can visualize this entire process using the powerful metaphor of a ​​fitness landscape​​. For many simple selection scenarios, we can imagine the population "climbing" a hill towards a state of higher mean fitness.

In the case of negative frequency-dependence, the landscape is a single hill. The peak of the hill corresponds to the stable mixture of types, and the population reliably climbs to this summit, preserving diversity.

But for positive frequency-dependence, the landscape is inverted. It's a ​​fitness valley​​. The lowest point of the valley is at the unstable equilibrium, $p^* = \frac{1}{2}$. The two "peaks," or states of maximum mean fitness, are at the boundaries: all-A ($p=1$) and all-B ($p=0$). Natural selection will always drive the population "uphill" on this landscape. Starting near the bottom of the valley, the population will inevitably roll up one side or the other, ending at one of the two monomorphic peaks. The population actively flees from the state of coexistence.

Positive frequency-dependent selection, from assortative mating (it's easier to find a similar mate if your type is common to Müllerian mimicry (different toxic species converging on the same warning signal to share the cost of educating predators), seems to be a potent force for purging diversity within a population.

But what if we zoom out? Imagine a landscape dotted with many smaller, isolated populations. In one patch, type A might, by chance, start out with a frequency of $0.6$. It will be driven to fixation. In another patch, it might start at $0.3$ and be eliminated. The result could be a regional mosaic: a collection of patches, each uniform in itself, but different from its neighbors. In this way, positive frequency-dependent selection, while destroying polymorphism locally, can paradoxically be a mechanism for generating and maintaining diversity on a grander geographical scale. It shows how the same evolutionary process can have vastly different consequences depending on the scale at which we look—a beautiful reminder of the interconnected complexity of the living world.

Having unraveled the principles of positive frequency-dependent selection, we are like a physicist who has just been handed a new fundamental law. At first, it explains a specific, curious phenomenon, but soon, we begin to see its reflection everywhere we look. This is the true beauty of a deep scientific idea—it doesn't just solve one puzzle; it provides a new lens through which to view the world. The principle that "it pays to be common" is such a lens. It is an engine of standardization, a force for convention, and its influence extends from the patterns on a butterfly's wing to the very language of life inscribed in our DNA. Let's embark on a journey to see just how far its echoes reach.

Perhaps the most vivid and classic illustration of positive frequency-dependent selection (PFDS) is found in the vibrant world of aposematism, or warning coloration. Imagine you are a young, naive bird hunting for insects and small animals. You see a brightly colored frog and, driven by hunger, you eat it. To your great regret, the frog is toxic, and you become violently ill. You have just learned a valuable lesson: "do not eat bright frogs."

Now, what if there are ten different species of toxic frogs in your forest, each with its own unique, fancy pattern? Your learning task just became much harder. You would have to sample, and get sick from, each of the ten patterns to learn to avoid them all. From the frogs' perspective, this is a disaster. Each species has to "pay the price" of educating generations of predators, a price paid in the lives of its members.

Here, PFDS provides a brilliant solution: convergence. Natural selection will strongly favor any frog that, by random mutation, happens to look a little more like an already common, established warning pattern. Why? Because by joining a larger, existing "club" of well-advertised toxicity, the individual frog reduces its personal risk of being the unlucky "learning meal" for a naive predator. The predators have already been educated by the vast numbers of the common type. It is the ultimate "safety in numbers" policy, written by evolution.

This phenomenon, known as Müllerian mimicry, gives rise to breathtaking "mimicry rings" in nature, where dozens of unrelated, well-defended species—like the venomous coral snakes of the Americas or the famous Heliconius butterflies of the Amazon—all converge on a single, shared uniform. This contrasts sharply with Batesian mimicry, where a harmless species copies a dangerous one. In that case, the mimic's fitness is subject to negative frequency-dependent selection; the more common the harmless fake, the more it dilutes the warning signal, and the more likely a predator is to call the bluff. In Müllerian mimicry, every member is an honest advertiser, and the more who join the chorus, the stronger and more effective the signal becomes for all.

The logic of "safety in numbers" comes with a fascinating wrinkle: a new, alternative warning pattern cannot easily get started. A single mutant with a novel pattern is, by definition, exquisitely rare. To a predator, it is not part of an established "brand" of danger; it is simply a weird-looking, and possibly delicious, novelty. It will likely be eaten before it has a chance to reproduce. This means there is a "hump" to get over. For a new pattern to spread, it must first reach a critical threshold frequency, a tipping point, perhaps by chance in a small, isolated population. Only once it is common enough can the reinforcing power of PFDS take over and drive it to fixation.

This idea of a tipping point helps us understand the stunning geographic diversity of mimicry rings. If two populations of the same species are separated by a mountain range or a wide river, they might, by chance, overcome the threshold for two completely different warning patterns. Each pattern is perfectly effective in its own local region. This creates a geographic mosaic of "local conventions," where the "correct" warning signal changes from one valley to the next. The strength of migration between these regions is a critical factor. If migration is low, local conventions can persist. If migration becomes too high, the homogenizing flow of individuals will likely cause one convention to overwhelm and replace the other, establishing a new, broader standard. The evolutionary path taken also depends on the starting conditions. If a community already has a dominant, high-frequency signal, a new species is most likely to evolve to match it, a process called "advergence." If several signals exist at comparable frequencies, they may all shift and "converge" toward a new, intermediate pattern.

This notion of a "convention" is the key to unlocking the true universality of PFDS. The principle is not really about color or predators; it's about the benefits of coordination. Think of the rule for which side of the road to drive on. There is no inherently "fitter" side, but there is an enormous advantage to everyone agreeing on one. This is a "coordination game," and its dynamics are governed by PFDS. Any individual who unilaterally decides to drive on the "wrong" side will quickly be selected against!

We see these dynamics playing out in the evolution of animal cultures. In many species of songbirds, males learn their songs from older males, and females often prefer to mate with males who sing the most common, "fashionable" local dialect. This preference creates PFDS: singing the hit song gets you more mates. Of course, the system is rarely perfect. There is always a low rate of "innovation" or error in song-learning, which constantly introduces new, rare dialects. The result is an equilibrium where the local dialect is highly common, but never completely fixed, in a beautiful balance between the pressure to conform and the trickle of novelty.

The logic of coordination extends to cooperation at the cellular level. Consider a partnership between a plant and a symbiotic fungus living in its roots. For the mutualism to work, their biochemistries must be compatible—they must "speak the same language." Imagine two types of plants ($P_1, P_2$) and two types of fungi ($F_1, F_2$) where the partnership is only beneficial if they match ($P_1$ with $F_1$, or $P_2$ with $F_2$). This is another coordination game. PFDS will drive the coevolving populations into one of two stable states: either the entire community becomes type 1, or the entire community becomes type 2. In this way, PFDS can structure entire ecosystems, partitioning species into distinct, internally compatible guilds or modules.

We now arrive at the most profound application of this principle, one that touches the very foundation of biology. A central mystery of life is the universality of the genetic code. With a few minor exceptions, every living thing on this planet, from a bacterium to a blue whale, reads the language of DNA in the exact same way. The codon G-G-U means "glycine" in you, in a mushroom, and in the virus that gives you a cold. Why this extraordinary uniformity? Out of the countless possible genetic codes, why did life settle on just one?

The answer, it seems, is a story of positive frequency-dependent selection on a planetary scale. In the early days of life, in the "primordial soup," different lineages of primitive cells likely had different, competing genetic codes. These early life-forms also engaged in rampant Horizontal Gene Transfer (HGT), swapping genes back and forth. Here lies the coordination game. If a cell received a gene from a donor with a different genetic code, that gene would be translated into a useless, garbled protein. It would be junk. But if it received a gene from a donor with the same code, it could be a wondrous new adaptation—the ability to metabolize a new sugar, or resist a toxin.

This created an immense fitness advantage for any lineage using the code that was, by chance, becoming more common. They had access to a larger "open-source community" of genetic innovations. The selective pressure was enormous. Once any single code reached a critical frequency—a global tipping point—it was a winner-take-all scenario. PFDS would have rapidly driven this code to near-universal dominance, locking all of life into a single, shared convention. Any lineage that subsequently "innovated" a new code would be instantly isolated, cut off from the global gene pool, and swiftly outcompeted.

So, the next time you marvel at a feat of genetic engineering—inserting a jellyfish gene into a fish to make it glow—remember that this is only possible because of a coordination game played billions of years ago. The unity of the genetic code is a frozen accident, an ancient convention established by the inexorable logic of positive frequency-dependent selection.

The drive to conformity, powered by PFDS, has consequences that ripple throughout the genome. When a successful new warning pattern or another favored convention sweeps through a population, it doesn't travel alone. Neutral genes that happen to be physically linked to it on the chromosome can be dragged along for the ride, a process known as genetic hitchhiking. In this way, PFDS can reshape vast landscapes of entire chromosomes, purging variation in regions surrounding the site of selection. From a single signal to the entire genetic operating system, PFDS is a powerful and unifying force, ensuring that sometimes, in evolution, it is indeed very, very good to be common.