Effective Population Size (Ne): Why Genetic Populations Shrink Below Headcount

What Ne Is and Why It Differs From the Headcount

Effective population size (Ne) is the size of a Wright–Fisher idealized population — one with random mating, non-overlapping generations, constant size, equal sex ratio, and Poisson-distributed offspring number — that would experience genetic drift, inbreeding, or loss of heterozygosity at the same rate as the real population under study. Because no real population meets the idealized assumptions, Ne is almost always smaller than the census size N.

The gap matters because the fundamental equations of population genetics contain Ne, not N. Each generation an idealized population loses a fraction 1/(2Ne) of its heterozygosity to drift, and inbreeding accumulates at the same rate. The variance in allele-frequency change per generation is p(1−p)/(2Ne).

• N (census size): everyone you can count, including juveniles and non-breeders.
• Ne (effective size): the genetically relevant number of breeders.

The ratio Ne/N is typically 0.1–0.5 in wild animal populations, and far smaller — around 10⁻⁶ — for the human species across its deep evolutionary history.

The Mechanism: How Each Factor Shrinks Ne, Step by Step

Ne falls below N whenever any idealized assumption is violated. The reduction operates through three dominant channels, each with a governing equation from Wright's foundational work:

• Unequal sex ratio (Wright 1931): only breeding individuals of both sexes transmit alleles. Ne = 4·Nm·Nf/(Nm + Nf), where Nm and Nf are breeding males and females. If one sex is rare — a harem-mating elephant seal, say — Ne is dragged toward that rare sex. Ten breeding males and 990 females give Ne ≈ 40, not 1000.
• Fluctuating size (Wright 1938): across t generations, the long-term Ne is the harmonic mean of the per-generation sizes, Ne = t / Σ(1/Nt). The harmonic mean is dominated by the smallest term, so a single bottleneck generation depresses Ne for many generations afterward.
• Reproductive-success variance: Ne ≈ (4N − 2)/(Vk + 2), where Vk is the variance in family size. Under Poisson reproduction Vk ≈ 2 and Ne ≈ N; when a few individuals dominate breeding, Vk climbs and Ne plummets.

Each mechanism amplifies random sampling of alleles between generations — the essence of genetic drift.

Key Quantities: Human Ne, Bottlenecks, and Concrete Numbers

The most striking illustration is our own species. Despite a census size near 8 billion, the long-term human effective population size inferred from genetic diversity is only about 10,000. This reflects that Ne integrates over the deep past — including severe bottlenecks 10,000–200,000 years ago and the Out-of-Africa founder events.

• Recent linkage-disequilibrium estimates (Tenesa et al. 2007): Ne ≈ 7,500 for the Yoruba (YRI) and ≈ 3,100 for European (CEU) and East Asian (JPT, HCB) samples — the lower non-African values reflect out-of-Africa bottlenecks.
• Mitochondrial DNA & Alu insertions: independently converge on Ne ≈ 10,000.
• Deep ancestral Ne: maximum-likelihood estimates suggest ~18,500 before ~1.2 Mya, dropping to ~8,500 afterward.

Wild examples: many threatened vertebrates have Ne in the low hundreds even when N is in the thousands, because Ne/N ≈ 0.1–0.25 is common. Cheetahs, having passed through a severe Pleistocene bottleneck, show genome-wide homozygosity consistent with a tiny historical Ne.

How Ne Is Measured: Genetic and Demographic Estimators

Because Ne is defined by a rate (of drift, inbreeding, or coalescence), it can be estimated several distinct ways, and each defines a slightly different flavor of Ne:

• Inbreeding Ne: from the rate of increase in homozygosity / inbreeding coefficient F per generation.
• Variance Ne: from the temporal variance in allele frequencies between sampled generations (the temporal method, Nei & Tajima 1981).
• Coalescent Ne: from present-day nucleotide diversity, since expected heterozygosity θ = 4·Ne·μ for diploids (μ = mutation rate per site per generation).
• Linkage-disequilibrium (LD) Ne: from the amount of non-random association between loci; drift generates LD inversely proportional to Ne. Notably, LD-based estimates are largely unaffected by natural selection (PLoS Genetics 2021).
• Coalescent HMM methods (PSMC/MSMC, Li & Durbin 2011): reconstruct Ne trajectories through time from a single diploid genome.

For a species like humans, θ ≈ 0.001 per bp and μ ≈ 1.2×10⁻⁸ per site per generation yield Ne on the order of 10⁴ — matching the other estimators.

Ne Versus Related Concepts: Census Size, Ne/N, and Effective Number of Breeders

Ne is easily confused with several neighbors; distinguishing them prevents serious errors in conservation planning:

• Census size N: the raw count. N tells you about ecology and extinction from demographic/environmental stochasticity; Ne tells you about genetics.
• Effective number of breeders (Nb): the effective size per reproductive cycle or cohort, distinct from the per-generation Ne; Nb and Ne differ predictably with generation length.
• Metapopulation Ne: subdivision plus limited migration can, counterintuitively, raise whole-species Ne by protecting alleles in separate demes (a consequence of the Wahlund effect and structured coalescence).
• Inbreeding vs. variance Ne: identical in a stable population but diverge sharply when size is changing — variance Ne looks forward (offspring sampling), inbreeding Ne looks backward (parent sampling).

The Ne/N ratio is the practical bridge: multiplying a monitored census by an empirically derived Ne/N (often 0.1–0.3) gives managers a usable genetic effective size without full genetic sampling.

Why Ne Matters: Conservation, Disease, and Open Questions

Ne is the number that decides whether a population's genome is safe. Small Ne means strong drift, rapid inbreeding, fixation of mildly deleterious mutations, and loss of adaptive potential.

• Conservation genetics: Franklin and Soulé's influential 50/500 rule holds that Ne ≥ 50 avoids acute inbreeding depression in the short term, while Ne ≥ 500 (many now argue ≥ 1,000) is needed to retain long-term evolvability by balancing mutation input against drift. The IUCN uses Ne thresholds in extinction-risk assessment.
• Selection efficiency: selection overrides drift only when the selection coefficient s exceeds ~1/(2Ne). In small populations, weakly deleterious alleles behave as effectively neutral and can fix — a driver of mutational meltdown.
• Human disease and biobanks: founder populations (Ashkenazi Jews, Finns, French Canadians) with small historical Ne carry elevated frequencies of specific recessive disease alleles, aiding gene mapping.

Open questions include how selection at linked sites (linked selection / genetic draft) systematically depresses Ne, why Ne/N varies so widely across taxa, and how to estimate contemporary Ne accurately for highly fecund marine species with enormous reproductive variance.

Factors that push effective size (Ne) below census size (N), with representative Ne/N effects

Factor	Governing relation	Effect on Ne	Example / typical value
Unequal sex ratio	Ne = 4·Nm·Nf / (Nm + Nf)	Ne collapses toward the rarer sex	10 males, 990 females → Ne ≈ 40, not 1000
Fluctuating size over generations	Ne = harmonic mean of Nt	Dominated by the smallest bottleneck	N = 1000, 1000, 10, 1000, 1000 → Ne ≈ 48
Variance in reproductive success	Ne ≈ (4N − 2) / (Vk + 2)	High variance in offspring number lowers Ne	Vk = 6 → Ne ≈ ½ N
Overlapping generations / age structure	Uses generation time & survivorship	Usually modest reduction	Ne/N ≈ 0.25–0.75 in many vertebrates
Population subdivision (Wahlund)	Depends on migration rate m, FST	Can raise metapopulation Ne	Structured demes retain more diversity
Human species (empirical)	Coalescent / LD estimators	Ne ≪ N by ~six orders of magnitude	Long-term Ne ≈ 10,000; N ≈ 8×10⁹