Effective Population Size in Genetic Drift — Epoche B2
Beyond Census: Re-evaluating Effective Population Size in Genetic Drift In population genetics, a foundational concept is that the strength of genetic drift [1] , the random fluctuation of allele frequencies, is inversely proportional to the population size. This implies that smaller populations experience stronger drift, leading to a faster loss of genetic diversity. While this intuition holds true in simplified theoretical models, it often overlooks crucial biological realities. The observed census population size ($N$), which is a simple count of individuals, frequently differs significantly from the genetically effective population size ($N_e$). The effective population size represents the number of individuals in an idealised population that would experience the same amount of genetic drift as the real [2] , non-ideal population. This discrepancy has profound implications for predicting genetic diversity changes, particularly in conservation efforts for endangered species. Genetic Drift in an Ideal Population Genetic drift is a stochastic process inherent to all finite populations. It describes the random changes in allele frequencies from one generation to the next, arising purely from chance in the sampling of gametes that form the next generation. Unlike selection, which acts on fitness differences, genetic drift is non-directional and can lead to the loss of advantageous alleles or the fixation of deleterious ones, especially in small populations. Over time, drift reduces genetic variation within a population and increases genetic differentiation between populations. To quantify genetic drift, population geneticists often refer to the Wright-Fisher model [3] , an idealised theoretical framework. This model describes a diploid, sexually reproducing population with discrete, non-overlapping generations. Key assumptions of the ideal Wright-Fisher population include: A constant population size of $N$ adult individuals. Random mating among all individuals. No selection, mutation, or migration. Each individual contributes gametes to a large, common gene pool, from which $2N$ gametes are randomly sampled to form the next generation's $N$ diploid individuals. Under these ideal conditions, the allele frequencies in the next generation are a random sample from the current generation. For a locus with two alleles, A and a, with frequencies $p$ and $q=1-p$ respectively, the expected change in allele frequency due to drift is zero. However, there is a variance associated with this change. The variance in the allele frequency $p'$ in the next generation is given by: $$ \text{Var}(p') = \frac{p(1-p)}{2N} $$ Here, $p$ is the frequency of an allele in the current generation, $1-p$ is the frequency of the alternative allele, and $N$ is the census population size. This equation quantifies the magnitude of genetic drift: a smaller $N$ leads to a larger variance in allele frequency, meaning stronger drift. The concept of effective population size ($N_e$) was developed to bridge the gap between this ideal model and real biological populations. We define $N_e$ as the size of an ideal Wright-Fisher population that would experience the same amount of genetic drift as the actual population under consideration. In other words, if a real population has a variance in allele frequency change of $\text{Var}(p')_{\text{real}}$, its effective population size $N_e$ is the value that satisfies: $$ N_e = \frac{p(1-p)}{2 \text{Var}(p')_{\text{real}}} $$ This definition allows us to account for various demographic factors that cause real populations to deviate from the ideal, leading to $N_e$ often being considerably smaller than $N$. We now explore two primary factors: unequal sex ratios and variance in offspring number. Unequal Sex Ratios and Effective Population Size One common demographic factor that reduces $N_e$ is an unequal sex ratio among breeding individuals. In many species, particularly those with polygynous or polyandrous mating systems, a small number of individuals of one sex may contribute disproportionately to the next generation's gene pool. For example, in a population where only a few dominant males mate with many females, the genetic diversity passed on to the next generation is effectively funnelled through these few males. This creates a genetic bottleneck, limiting the overall genetic contribution and increasing the impact of drift. Consider a diploid population consisting of $N_m$ breeding males and $N_f$ breeding females. The total census population size is $N = N_m + N_f$. The effective population size for such a population is given by: $$ N_e = \frac{4 N_m N_f}{N_m + N_f} $$ This formula highlights that $N_e$ is maximised when $N_m = N_f$, meaning an equal contribution from both sexes. If the number of breeding males and females is perfectly balanced (e.g., $N_m = N_f = N/2$), then $N_e = \frac{4 (N/2)(N/2)}{N/2 + N/2} = \frac{4 N^2/4}{N} = N$. However, as the sex ratio becomes increasingly skewed, $N_e$ drops dramatically. For instance, imagine a population of $N=100$ individuals. If there are 50 breeding males and 50 breeding females ($N_m=50, N_f=50$), then $N_e = \frac{4 \cdot 50 \cdot 50}{50+50} = \frac{10000}{100} = 100$. In this balanced scenario, $N_e$ equals $N$. Now consider a skewed ratio, such as 10 breeding males and 90 breeding females ($N_m=10, N_f=90$). The effective population size becomes $N_e = \frac{4 \cdot 10 \cdot 90}{10+90} = \frac{3600}{100} = 36$. Despite having 100 individuals, this population experiences genetic drift as if it were a much smaller ideal population of only 36 individuals. If the skew is even more extreme, say 1 breeding male and 99 breeding females ($N_m=1, N_f=99$), then $N_e = \frac{4 \cdot 1 \cdot 99}{1+99} = \frac{396}{100} = 3.96 \approx 4$. In this case, $N_e$ is less than 4% of the census size, indicating extremely strong genetic drift. The chart shows three scenarios: a 50:50 male-to-female ratio, a 10:90 ratio, and a 1:99 ratio. For each