Population Genetics

Population genetics is the mathematical and empirical study of how allele frequencies change in populations over time under the forces of natural selection, genetic drift, mutation, and gene flow. It is the quantitative backbone of evolutionary biology — the field that transformed Darwin's qualitative account of descent with modification into a body of predictive theory amenable to mathematical analysis and experimental test. Without population genetics, evolutionary biology has mechanisms but no equations; with it, mechanisms become models that generate predictions, and predictions can be wrong.

The field has its roots in the early twentieth-century work of Fisher, Wright, and Haldane, who demonstrated that Mendelian genetics and Darwinian selection were not only compatible but mutually reinforcing. The Modern Synthesis of the 1930s–1940s was, in substantial part, a synthesis achieved by population genetics.

Core Theory

The Hardy-Weinberg equilibrium is the null model: in an infinitely large, randomly mating population with no mutation, selection, migration, or drift, allele frequencies remain constant indefinitely, and genotype frequencies take predictable proportions. No real population satisfies these conditions. That is the point. Hardy-Weinberg is useful precisely as a baseline from which deviations can be measured and attributed to specific forces.

Effective population size (N_e) is the size of an idealized Hardy-Weinberg population that would exhibit the same rate of genetic drift as the actual population. It is almost always smaller than the census population size, sometimes dramatically so. The ratio of N_e to census size reflects fluctuating population sizes, unequal sex ratios, variance in reproductive success, and geographic structure. N_e is the key parameter determining whether selection or drift dominates: when the product of N_e and the selection coefficient (N_e × s) is much less than 1, drift overwhelms selection; when much greater than 1, selection dominates. This is the framework underlying the Neutral Theory of Molecular Evolution and the nearly neutral theory.

Selection coefficients measure the relative fitness advantage or disadvantage of a genotype. A selection coefficient of 0.01 means the genotype leaves 1% more descendants per generation than the reference genotype. Selection this weak is nearly invisible in the short run but cumulatively powerful over thousands of generations — and nearly invisible to natural selection in small populations, where drift can override it. The tension between selection strength and population size is central to understanding which variants spread and which are lost.

Coalescent Theory

Coalescent theory, developed by John Kingman in 1982, models the genealogical history of a sample of gene copies — working backward in time to find the common ancestor of the sample. It transformed population genetics from a forward-time (how will frequencies change?) discipline to one that could make inferences about past history from present genetic data. Modern genomics applies coalescent-based methods to infer historical population sizes, migration patterns, selection events, and demographic histories from genome-wide data.

The coalescent revealed that the genealogical history of a sample is far more structured than random: lineages coalesce faster when populations are small, and genetic diversity is a direct function of N_e × mutation rate. Human genome-wide diversity estimates consistently imply an effective ancestral population size of roughly 10,000 — not because humans are descended from only 10,000 individuals, but because the genealogical bottlenecks in human prehistory reduced N_e to this effective value. This is a population-genetic fact, not a theological one, and it constrains the space of historical scenarios compatible with genomic data.

Limits of the Adaptationist Reading

Population genetics has been appropriated — and distorted — by the adaptationist program that assumes all genetic variation is maintained by selection. The evidence contradicts this at the molecular level: most variants in the human genome have tiny or zero fitness effects, and their frequency distributions match neutral or nearly neutral models better than strongly selective ones.

The implication is uncomfortable for adaptationist narratives of human genetic variation: most differences between human populations are not adaptive. They are the product of drift and historical accident — founder effects, bottlenecks, migration routes. The field of genomics regularly finds candidate genes for population differences and interprets them as adaptive without adequately testing the neutral null hypothesis. This is a methodological failure that population genetics has the tools to correct, if researchers use those tools rather than generating just-so stories.

Population genetics is a humbling discipline. It demonstrates that most of evolution is not the story of organisms heroically adapting to environments. It is the story of variants drifting through populations like flotsam on a current — some surviving, most not, none of it planned.