Population genetics - Evolutionary Forces and Advanced Dynamics

Evolutionary Processes in Population Genetics Introduction Evolution occurs when allele frequencies change over time within populations. While we often think of evolution as organisms becoming better adapted, the actual mechanisms driving change are more subtle. Four main evolutionary forces shape allele frequencies: selection, mutation, genetic drift, and gene flow. Each works through different mechanisms, and they interact in complex ways depending on population size and the strength of each force. Understanding how these forces work separately and together is essential for predicting how populations will evolve. Selection Selection is the process where certain traits increase or decrease in frequency because they affect survival or reproduction. Natural selection acts when individuals carrying different alleles have different chances of surviving and reproducing. Sexual selection is a special case where traits increase in frequency because they improve mating success, even if they reduce survival. Fitness and Selection Coefficients Fitness describes the reproductive success of an individual or genotype. Rather than using absolute fitness values, population geneticists typically use relative fitness, which compares fitness values to the most successful genotype. The fitness of a genotype is expressed as: $$w = 1 - s$$ where $s$ is the selection coefficient. This coefficient represents the reduction in fitness relative to the most fit genotype. For example, if $s = 0.1$, that genotype has 10% reduced fitness compared to the optimum. An $s$ value of 0 means no selection (neutral), while $s = 1$ means the genotype is lethal. Key insight: Selection converts differences in fitness into changes in allele frequency across generations. An allele that increases fitness will tend to increase in frequency over time, while one that reduces fitness will tend to decrease. When Selection Overcomes Drift A critical principle in population genetics is understanding when selection is strong enough to change allele frequencies despite the random changes caused by population sampling. This occurs when: $$s > \frac{1}{Ne}$$ where $Ne$ is the effective population size (the number of individuals in an idealized population that would experience the same amount of genetic drift as the real population being studied). What this means: If selection is stronger than the random effects of population size, selection will win and allele frequencies will change predictably. If selection is weaker than drift, random sampling will dominate and allele frequencies may change randomly regardless of fitness differences. Fixation Probability of Beneficial Mutations When a new, beneficial mutation appears in a population, it will not necessarily spread to fixation (become present in every individual). It might be lost by chance even though it improves fitness. The probability that a beneficial allele eventually becomes fixed is approximately: $$P{\text{fix}} \approx 2s$$ Why only 2s? Even though the mutation is beneficial, drift can eliminate it before selection has time to increase it substantially. This probability assumes the mutation is rare when it arises. The stronger the selection coefficient $s$, the more likely the mutation will spread to fixation. Dominance: How Heterozygotes Matter The fitness effects of an allele depend critically on whether it's paired with itself (homozygous) or with a different allele (heterozygous). This relationship is described by dominance. The dominance coefficient $h$ describes the fitness of a heterozygote relative to the two homozygotes. Specifically: Fitness of homozygote with harmful allele: $1 - s$ Fitness of heterozygote: $1 - hs$ Fitness of homozygote with beneficial allele: $1$ When $h = 0$, the allele is completely recessive (no fitness effect in heterozygotes). When $h = 0.5$, fitness effects are additive. When $h = 1$, the allele is completely dominant. Why this matters for evolution: A recessive deleterious allele (small $h$) can hide in heterozygotes and persist in populations even when harmful. A dominant beneficial allele spreads quickly because every copy contributes to fitness, even in heterozygotes. Epistasis: When Loci Interact So far we've considered selection at a single locus in isolation. But fitness is rarely determined by just one gene. Epistasis occurs when the fitness effect of an allele at one locus depends on which alleles are present at other loci. Diminishing-returns epistasis describes a common pattern: beneficial mutations provide smaller fitness increases when the organism already has high fitness from other mutations. For example, if you already have several mutations that improve enzyme function, adding another similar beneficial mutation helps less than if you had no prior mutations. Synergistic epistasis refers to cases where deleterious mutations have more harmful combined effects than expected. If two mutations each reduce fitness by 10%, you might expect combined damage of 20%, but synergistic epistasis produces worse effects. Practical consequence: Epistasis means that the fitness benefit of a mutation depends on the genetic background—the suite of other alleles present. This makes prediction of evolution more complicated in real organisms. Mutation Mutation creates heritable changes in DNA. These range from single nucleotide changes (point mutations) to large-scale rearrangements like duplications or deletions. Without mutation, populations would only shuffle existing genetic variation. Mutation is the ultimate source of all new alleles. Mutation Rates and Bias Mutations don't occur at random. Different types of mutations have different rates, and these differences, called mutation bias, can systematically influence evolution. For example, transitions (purine↔purine or pyrimidine↔pyrimidine substitutions) occur more frequently than transversions. Some organisms show bias toward AT or GC base composition. These biases exist because of the chemical properties of DNA and how DNA polymerase makes errors. Mutation bias can direct adaptation: When populations are large and asexual (undergoing clonal reproduction), multiple mutations spread through different lineages simultaneously. Mutation types that occur more frequently spread more easily, so mutation bias can favor certain adaptive paths over others, even when multiple mutations could improve fitness similarly. Mutation-Selection Balance When a deleterious mutation is introduced, selection works to remove it, but mutation keeps reintroducing it. At equilibrium, these forces balance. The deterministic mutation-selection balance equation gives the equilibrium allele frequency: $$f = \frac{u}{s}$$ where $u$ is the mutation rate and $s$ is the selection coefficient against the allele. Interpretation: Higher mutation rates or weaker selection leads to higher equilibrium frequencies of deleterious alleles. This is why mutation-selection balance predicts more genetic load in organisms with high mutation rates or small populations (where $s$ is effectively small). Genetic Drift While selection and mutation are deterministic forces (they produce predictable changes), genetic drift is random. Genetic drift occurs because parents don't pass on all their alleles equally to offspring—it's random sampling. In a population where 50% of alleles are type A and 50% are type a, by chance the next generation might inherit 51% A and 49% a, just from random sampling. How Drift Works Imagine an allele present in exactly 50% of parents. The offspring generation should have 50% of this allele on average—but the actual percentage could be 48% or 52% or 55% just by chance. This randomness causes allele frequencies to wander up and down unpredictably over generations. Key consequence: Even with no selection and no mutation, allele frequencies change. Eventually, drift causes alleles to reach fixation (frequency = 1) or loss (frequency = 0). This means populations lose genetic variation over time purely from drift. Drift Strength Depends on Population Size The smaller the population, the stronger the random sampling effects, and the faster allele frequencies change randomly. In large populations, random sampling is less pronounced. The strength of drift is inversely related to effective population size: drift is severe in small $Ne$ and weak in large $Ne$. Drift, Selection, and Population Size The interplay between drift and selection is crucial. In very small populations, drift dominates, and even moderately deleterious mutations can reach fixation. In large populations, selection dominates, and even slightly beneficial mutations can spread. The boundary occurs roughly where $s \approx 1/Ne$—when selection and drift have similar strengths. This has major implications: large organisms with small populations (like elephants) experience strong drift. Microorganisms with enormous populations experience weak drift and strong selection. This is why bacteria can evolve rapidly even to weak selective pressures. Gene Flow Gene flow is the movement of alleles between populations via migration. When individuals move from one population to another and breed, they introduce new alleles and change allele frequencies in both populations. Effects of Gene Flow Gene flow can: Introduce beneficial alleles into populations that lack them Introduce deleterious alleles that cause a migration load (reduced population fitness) Counteract inbreeding depression by introducing genetic variation Break down genetic differentiation between populations, eventually homogenizing them High gene flow between populations tends to make them genetically similar, while low gene flow allows them to diverge as different mutations arise and drift occurs independently in each. <extrainfo> Horizontal Gene Transfer In prokaryotes, horizontal gene transfer (HGT) moves genetic material between unrelated organisms, not through reproduction but through mechanisms like bacterial conjugation or viral transduction. This allows bacteria to acquire useful genes (like antibiotic resistance) from distantly related species. HGT has also been important in eukaryotic evolution—evidence suggests that mitochondria and chloroplasts were originally free-living bacteria that were acquired through ancient HGT events, fundamentally reshaping the evolution of eukaryotic cells. </extrainfo> Linkage Disequilibrium So far, we've often treated alleles at different loci as independent. But this isn't always true. Linkage disequilibrium (LD) occurs when alleles at different loci are non-randomly associated—certain combinations occur more frequently than expected by chance. Why Linkage Disequilibrium Exists and Breaks Down Linkage disequilibrium arises when: Two loci are physically close on a chromosome, so recombination rarely separates them A new mutation arises and hasn't had time to recombine away Populations were recently admixed Recombination breaks down LD, but slowly. The closer two loci are, the less recombination occurs between them, so LD persists longer. Genetic Hitchhiking When selection favors an allele at one locus, linked alleles at nearby loci are carried along, even if they're neutral or harmful. This is genetic hitchhiking or genetic draft. A neutral allele linked to a beneficial mutation that's spreading will increase in frequency alongside it. When the beneficial mutation reaches fixation, the hitchhiking allele reaches fixation too—not because of selection, but because it was linked. The Hill–Robertson effect describes how linkage interferes with selection. Multiple beneficial mutations competing in a small region cannot be optimally selected because recombination can't efficiently combine them all. This slows adaptation. Background selection is the opposite problem: deleterious mutations linked to beneficial ones can persist temporarily because they're in the same chromosome region as a beneficial mutation that's spreading. Recombination eventually separates them, but this takes time. How Mutation, Selection, and Drift Interact The outlined content emphasizes a crucial modern insight: mutation, selection, and drift don't operate independently. Their interaction shapes evolution in ways that none could produce alone. The Evolution of Mutation Rates Themselves Mutation rates are not fixed—they evolve. Natural selection favors lower mutation rates because high mutation rates damage organisms. However, in finite populations, genetic drift can fix higher-mutation-rate alleles simply by chance. This creates a balance: selection pushes mutation rates down, but drift prevents them from becoming arbitrarily low. The drift-barrier hypothesis explains why organisms can't achieve perfectly low mutation rates: selection against mutations that increase the mutation rate becomes too weak relative to drift when populations are finite. The absolute minimum mutation rate achievable is determined by the population size—large-population organisms can evolve lower mutation rates than small-population organisms. Empirical pattern: Organisms with large effective population sizes (like bacteria and some invertebrates) have lower per-generation mutation rates than organisms with small populations (like large mammals and plants with small population sizes). Clonal Interference and Mutation Bias in Adaptation In large asexual populations, adaptation depends heavily on which mutations arise. If beneficial mutations occur at different rates due to mutation bias, more-frequent mutations fix more easily, even if rare mutations would be more beneficial. This is clonal interference—multiple beneficial mutations arise in different lineages and compete, with the first to reach high frequency typically winning out. Mutation bias can therefore directly steer evolution toward certain adaptive solutions and away from others, independent of which is actually best. This means evolution isn't always "optimal"—it follows paths determined partly by random mutation patterns. Summary The four evolutionary forces—selection, mutation, genetic drift, and gene flow—constantly shape allele frequencies. Their relative importance depends on population size, mutation rates, and the strength of selection. In large populations with efficient selection and weak drift, evolution tracks fitness landscapes predictably. In small populations where drift dominates, evolution becomes increasingly random. Modern population genetics recognizes that these forces interact in complex ways: linked selection creates pseudo-drift, mutation bias constrains adaptive paths, and the evolution of mutation rates itself depends on balance between selection and drift. Understanding these interactions is essential for predicting how real populations will evolve.