intermediate 8 min read
Life sciences · Topic
Response to Selection
probability theory · gaussian distribution · optimization
The response to selection $R$ is the change in trait mean between the selected parents and their offspring. The Breeder's equation — the simplest and most powerful result in quantitative genetics — links response directly to heritability and the intensity of selection applied to the parental generation.

The Breeder’s Equation

\[R = h^2 S\]

where $S$ is the selection differential (the mean phenotype of selected parents minus the population mean) and $h^2$ is narrow-sense heritability. Equivalently, in terms of selection intensity $i$ (the standardised selection differential):

\[R = i\, h^2\, \sigma_P = i\, h\, \sigma_A\]

For a truncation selection scheme where the top fraction $p$ of a normal distribution is selected, $i = \phi(t)/p$ where $t$ is the truncation point and $\phi$ is the standard normal density.

Selection Differential and Intensity

Proportion selected ($p$) Selection intensity ($i$)
0.01 2.665
0.05 2.063
0.10 1.755
0.25 1.271
0.50 0.798

The Selection Index

When multiple traits are measured, the optimal linear index for selection on a single goal trait uses index coefficients $\mathbf{b}$:

\[I = \mathbf{b}^\top \mathbf{x}, \quad \mathbf{b} = \mathbf{P}^{-1}\mathbf{G}\mathbf{a}\]

where $\mathbf{P}$ is the phenotypic variance-covariance matrix, $\mathbf{G}$ is the genetic variance-covariance matrix, and $\mathbf{a}$ is a vector of economic weights. The index maximises the correlation between $I$ and the aggregate genotype $H = \mathbf{a}^\top \mathbf{g}$.

Long-Term Response and Limits

Sustained directional selection erodes additive genetic variance: $V_A$ declines as favourable alleles approach fixation. The Robertson–Hill limit estimates the total response from drift and selection in a finite population of size $N_e$:

\[R_{\infty} \approx 2\, N_e\, R_1\]

where $R_1$ is the single-generation response. In practice, mutation-selection balance and frequency-dependent effects sustain long-term response beyond this prediction.