Genetic drift Part 1

= 6 = 0.6! = 4 = 0.4! = 4 = 0.4! = 6 = 0.6! = 5 = 0.5! = 5 = 0.5! = 3 = 0.3! = 7 = 0.7! = 8 = 0.8! = 2 = 0.2! = 7 = 0.7! = 3 = 0.3! SAMPLES from population! Small samples are LESS likely to match with theoretical expectation! Equivalent to the theoretical expectation! in an infinite population! TRUE allele frequencies in the population! = 50 = 0.5! = 50 = 0.5! consider ENTIRE (ideal) population! Genetic drift (or random sampling error) can change allele frequencies! Random walks... describe fluctuations in the frequencies of alleles that are neutral with respect to fitness! demes ! q q q q Over generations, random sampling error occurs...! During each generation, the large number of zygotes produced is reduced (via mortality that is random with respect to fitness) to some number of breeding individuals! – allele frequencies change (drift) –! p = 0.47 A2 = g = 0.53 What happens to the frequencies of p and q over the long term...?! Alleles drift to fixation in each deme. ! Which allele gets fixed...?! The theory of genetic drift laid out by Sewell Wright during the 1930s and 1940s. Taken up by Motoo Kimura in the 1950s.! Natural selection & genetic drift are the 2 most important causes of evolutionary change in populations. ! The relative importance of each remains controversial.! Genetic drift as a null hypothesis... All alleles @ all loci are subject to genetic drift. The same cannot be said for natural selection.! Evolution by genetic drift...! (1)! Allele frequencies fluctuate at random within a population & eventually one (or another) allele will be fixed.! (2)! The magnitude of change is inversely related to the size of the sample; drift is stronger in smaller samples.! (3)! Drift is NOT repeatable. Different samples from identical populations will differ from one another.! Consequences of genetic drift...! (1)! Short term: random fluctuations lead to fixation of alleles! (2)! Long term: loss of GV, decline of heterozygosity, increase in the variance among populations! (3)! Evolutionary change occurs as the result of genetic drift, but genetic drift cannot produce adaptation! Probability of fixation! 1.0 0.8 (1)! Population size! Higher probability of fixation in smaller populations.! Frequency 0.6 of allele A1 0.4 0.2 In a diploid population, the average time to fixation for a novel, neutral allele that DOES become fixed is 4N generations.! 0.0 0 100 200 300 Generation N = 10! (40 generations)! N = 10,000! (40,000 generations)! … can be a LONG time if populations are large! 400 500 Probability of fixation! 1.0 0.8 (2)! Allele frequency! The probability that a particular allele will fix in a population is equal to its initial frequency in the population (Box 6.3).! initial number ! of copies of allele! <- total number! of alleles in ! population! Frequency 0.6 of allele A1 0.4 0.2 0.0 0 100 200 300 400 500 Generation At any time, a particular allele’s probability of fixation is equal to its frequency at that time.! z Within populations there is a decline in heterozygosity over time.! 1.0 0.8 heterozygosity in next generation! / Frequency 0.6 of genotype A1A2 0.4 0.2 0.0 0 100 200 300 400 500 heterozygosity in current generation! number individuals (N) in a population! Generation Allele A2 is lost l I Allele A1 is lost Heterozygosity Generation Empirical evidence for genetic drift Drosophila, •! •! •! Maintained 107 populations for 19 generations, keeping total Number of populations Frequency of bw 75 allele Number of populations rTfl KTk ^ 0 Frequency of bw 75 allele 1 -rffT-fTk 1 n-rfK 2 ^4 ^ n = Generation 11 Q n 12 ru 0 13 1 14 J] 15 3 - 4 ^ rrv 5 Q riTh 6 ^ o .00 . 13 . 25 . 38 . 50 .63 . 75 . 88 1.00 Frequency of bw 75 allele 16 30 n 10 0 .00 .13 .25 .38 .50 .63 .75 .88 1.00 Frequency of bw 75 allele 19 Generation 1 Number of populations Frequency of bw 75 allele 0.5 in 8CD >*, 0.4 - 3 CL C m CL c CD >; o 0.3 - N \ O o C > o O 0.2 - S> 01 - I -O CD cn 3 o.o 1 0 5 10 15 20 Generation Number of populations Frequency of bw 75 allele Differentiation with drift: a spatio-temporal genetic analysis of Galapagos mockingbird populations IMimus spp.) Paquita E. A . Hoeck 1 * , Jennifer L. Bollmer2’ , Patricia G . Parker2 ' and Lukas F. Keller 1 Phil. Trans. R. Soc. B (2010) 365, 1127-1138 Figure 1 . Distribution of the four mockingbird species in Galapagos with island names and locations from which contemporary (o) and historic (x) samples were obtained . On Isabela (AL and VL), St Cruz (PtA and GP), San Cristobal (N and S) and Espahola (MZ and PC) we collected samples in two different locations. Wolf D Af part'ulus OB OI &4 Ttifgg OB o AL ItahrU * O to)tni * St Cm/ a J ° or A l* VL M tri/asciatus M melunotis r 08 9 » S MS \ S*i M . mac donaldi o Historic! Contemporary! Heterozygosity! Allelic richness! \ / / \ \ / / / / / / / / \ \ 1.0 \ \ 4 \ \ \ \ / i \ \ \ \ i i ns 3 \ \ i i a \ \ \ \ i / / / \ \ / / 0.6 - L> * < 0.4 1 0.2 0 0 tUund < M/e ( In ) Figure 2 Expected heterozygosity (He) >nd allelxc richness AR) of contemporary and natural Ingamhm of island size ( in hah Tn tmprove visual representation, points dm overlapped ? funoJn Bf a* 0 were tightly moved (filled (a ) -I v Difference between historical and contemporary populations! X X X -2 £ . 'A k - 3 B a x X ' • X X x was GREATER for small compared to large populations! x X X X s & X X -4 s # '# . X X X -5 *. • X X X X X X X -6 T T ( h ) 10500 10000 9500 . / 2500 > z- / / 2000 . 1500 . 1000 500 0 Figure 3 (a) Temporal Fsr between contemporary and historic populations as a function of island size (in ha ; log scale) Differentiation clearly increased more strongly in smaller populations as shown by the dotted regression line (Fst 0.899 -0.354 * log(island size)) Overlapping data points from islands with similar size were slighdy modified to improve visual representation. (6) Estimates of effective population size (Ne) with lower and , where available, upper estimation limits plotted against island size (in ha; log scale). The dotted line shows the linear regression line (Nc 507.07 + 148.94 * log(island size)) A strong posi- MI =- =- island size ( In ) . . tive linear correlation was also found when both Ne and island size were In-transformed (r2 = 0.92, p < 0.0001). Natural history of genetic drift ! (1)! Small populations! e.g., !naturally small populations (ephemeral pools ! !or edaphic specialists)! !species introductions (founder effects) !