Genetic drift (or random sampling error) can change allele frequencies!

consider ENTIRE (ideal) !

= 50 = 0.5! = 50 = 0.5! SAMPLES from population! TRUE allele frequencies in the population! = 3 = 0.3! = 6 = 0.6! = 7 = 0.7! = 4 = 0.4! Equivalent to the theoretical expectation! in an infinite population! = 8 = 0.8! = 4 = 0.4! = 2 = 0.2! = 6 = 0.6! Small samples are LESS likely to = 7 = 0.7! = 5 = 0.5! match with theoretical expectation! = 3 = 0.3! = 5 = 0.5! Random walks... describe fluctuations in the frequencies of alleles that are neutral with respect to fitness! demes !

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) –! 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 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 . ! 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 .! 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 ! Probability of fixation! 1.0

0.8 (1)! ! Frequency 0.6 of Higher probability of fixation in allele A1 0.4 smaller populations.! 0.2 In a diploid population, the average time to fixation for a novel, neutral allele that 0.0 DOES become fixed is 4N generations.! 0 100 200 300 400 500 Generation

N = 10! N = 10,000! (40 generations)! (40,000 generations)!

… can be a LONG time if populations are large! Probability of fixation! 1.0


Frequency 0.6 of allele A1 (2)! Allele frequency! 0.4

The probability that a particular 0.2 allele will fix in a population is 0.0 equal to its initial frequency in 0 100 200 300 400 500 Generation the population (Box 6.3).!

initial number ! At any time, a particular allele’s of copies of allele! probability of fixation is equal to its total number! frequency at that time.! of alleles in ! population! Within populations there is a decline in heterozygosity over time.!

1.0 heterozygosity in next generation! 0.8

Frequency 0.6 of genotype A1A2 0.4

0.2 heterozygosity in number individuals 0 100 200 300 400 500 current generation! 0.0 (N) in a population! Generation

Allele A2 is lost Allele A1 is lost Heterozygosity

Generation Empirical evidence for genetic drift ! Drosophila, Buri (1956)!

•! 107 populations of flies (8 females & 8 males)!

•! All flies were heterozygotes (bw75/bw)!

•! Maintained 107 populations for 19 generations, keeping total size equal across generations.!


0 1

0 1

Generation 1! 107

0 1

19th generation!


0 1 50 km! Historic! Contemporary! Heterozygosity! Allelic richness! Difference between historical and contemporary populations!

was GREATER for small compared to large populations! Natural history of genetic drift ! (1)! Small populations! e.g., !naturally small populations (ephemeral pools ! !or edaphic specialists)!

! introductions (founder effects)!

Bufo punctatus! Founder effects - the loss of following the establishment of a new population by a few individuals! Natural populations! When we describe the genetic structure of natural populations, the data usually are not based on experimental manipulations, nor do we know the history of the populations, but...! we can infer the causes of evolution (e.g., genetic drift or selection) by interpreting the patterns of variation that we observe.!

Molecular variation @ 2 loci in Mus musculus! 42 populations of mice from widely separated barns in TX! !– small & large populations – ! Mean allele frequency & variance among populations! Natural history of genetic drift !

(2)! Effective population size < actual population size! e.g., !sexually immature individuals! !restrictive sexual systems! !skewed sex ratios! !population size fluctuations! What is population size?! !Ecological sense, count # individuals! What is population size?!

!Evolutionary sense, what matters is the number that contribute ! ! !genes to the next generation!

(= every individual reproduces)! effective population size, Ne! … the number of individuals in an ideal population in which the rate of genetic drift would be the same as it is in the actual population.!

1) !Equal numbers of males and females! 2) !No sexual or natural selection (i.e., all individuals have an equal probability of contributing gametes to next generation)! 3) !Population size remains the same in each generation! Effective population size, Ne!

Breeding structure of population! !number of sexually mature individuals! !unequal sex ratios!

Unequal variance in mating! 8N restrictive sexual systems that limit Ne " individuals from mating! Vm +Vf + 4

Population fluctuations in size! ! Natural history of genetic drift !

(1)! Small populations! e.g., !naturally small populations (ephemeral pools ! !or edaphic specialists)! !species introductions (founder effects)!

(2)! Effective population size < actual population size! e.g., !sexually immature individuals! !restrictive sexual systems! !skewed sex ratios! !population size fluctuations!

(3)! When historically large populations are reduced in size! e.g., !bottlenecks or fragmentation! Population bottlenecks - !the loss of genetic diversity following a !reduction in population size!

Northern elephant seal, Mirounga angustirostris!

•!historically distributed along Pacific coast (NA); > 150,000 seals! •! 1800’s hunted almost to (by 1890, < 20 individuals)! •!recovery efforts began 1900’s! •!current population ca. 100,000 individuals, but...! Hoelzel et al. (1993 J Heredity 84:443)! 41 allozyme loci, 67 individuals, 2 locations ! Bonnell & Selander (1974 Science 184:908)! 24 allozyme loci sampled in 159 individuals across 5 populations ! ~ all loci monomorphic ! Population bottlenecks - !the loss of genetic diversity following a !reduction in population size!

Southern elephant seal, Mirounga leonina!

Gales et al. (1989 Mar Mammal Sci 5:57)! 35 allozyme loci sampled! 196 individuals from 2 ! 25K


15K 1810-20! 1940! 1962! 1994!




1.! Controlled pollinations to assess the compatibility status for two African species, ferocissimum and L. pumilum

2.! Investigate S-RNase allelic diversity (present day: genotypes of individuals and long-term: trans-generic lineages)

3.! Explore patterns of at the S-RNase locus Molecular of tribe ()

4th largest group in family (~85 spp.)

•! Long-lived, woody shrubs •! Drought, salt tolerant •! Three genera in Lycieae •! Worldwide distribution

Lycium (~80 species) sampling ~88% of species 72 Lycieae (all 3 genera) 89 taxa total Nolana, Sclerophylax, Jaborosa (outgroups)

GBSSI (waxy) Aligned length = 1968 bp

ML analysis TVM + I -lnL = 6589.81 200 ML bootstrap replicates

Levin & Miller (2005) Am J Bot Levin et al. (2007) Acta Hort Biogeography A single dispersal event to the Old World Originated in So. America

Josh Shak ‘06

Dispersal to No. America

Julian Damashek ‘09

Subsequent movement between the Americas Self-incompatibility is widespread among angiosperms and is established as a mechanism to avoid .

Gametophytic S-RNase based SI

•! Two genes: pistil (S-RNase gene); pollen (SLF gene) When the haploid genotype of the pollen matches either of the two S-RNases expressed in the pistil of the maternal parent, pollen tube growth is terminated.

•! Heterozygosity of individuals •! High allelic diversity within populations & species •! Long-term persistence of alleles (i.e., trans-generic )

CLASSIC case of negative frequency dependence Species Trans-generic lineage (TGL) defined as the smallest phylogeny that includes an S-RNase from your group of interest and an S-RNase from another group

S-RNase gene genealogy Lycium!

30 MYA Petunia!




Interpretation is that the allelic lineage predates the Petunia! divergence of the genera (e.g., Solanum and Lycium) S-RNase gene (~370 bp)

9 S-RNases Lycium cestroides (C1-C9) 11 S-RNases Lycium (5 spp.) 10 S-RNases Solanaceae (6 spp.) Trans-generic lineage (TGL)

ML analysis defined as the smallest clade that TVM + I + G includes an S-RNase from Lycium

-lnL = 7063.09 cestroides & an S-RNase from another 50 ML bootstrap searches

Alleles are OLDER than the species in which they reside Petunia & Lycium ~ 40 MYA Lycium & Solanum ~ 30 MYA Species obtusa (Bianchi et al. 2001) phylogeny Lycium pallidum (Miller & Venable 2002)

GBSSI (~2000 bp)

ML analysis TVM + I Self-incompatibility has been documented across tribe Lycieae & is present in most -LnL = 6589.81 major in Lycium

Lycium californicum (JR Kohn, pers comm)

Lycium andersonii (Richman 2000)

Lycium berlandieri (Miller & Venable 2002)

Lycium parishii (Savage & Miller 2006)

Lycium cestroides (Aguilar & Bernardello 2001; Miller et al. in prep) Species Grabowskia obtusa (Bianchi et al. 2001) phylogeny Lycium pallidum (Miller & Venable 2002; Miller, unpublished data)

GBSSI (~2000 bp)

ML analysis TVM + I Studies of the S-RNase gene reveal high allelic diversity (e.g., sequence diversity, -LnL = 6589.81 numbers of trans-generic allelic lineages)

Lycium californicum (JR Kohn, pers comm)

Infer GSI in the ancestor of Lycium Lycium andersonii (Richman 2000)

Lycium berlandieri (Miller & Venable 2002)

Lycium parishii (Savage & Miller 2006)

Lycium cestroides (Aguilar & Bernardello 2001; Miller et al. in prep)

Controlled pollinations

174 SELF 163 CROSS 179 CONTROL n = 16

Lycium ferocissimum

165 SELF 136 CROSS 177 CONTROL n = 14 plants

Lycium pumilum Controlled pollinations

Cross Self Control

Both species are strongly self-incompatible •! Outcrossing resulted in a 51- (L. ferocissimum) or 93-fold (L. pumilum) increase in seeds per flower compared to self pollination. S-RNase sequence diversity S-RNase sequence diversity

24 unique S-RNase sequences •! ranged from 363–384 bp •! pairwise AA divergence = 0.34 •! 13/15 (87%) individuals heterozygous

Old World S-RNases were aligned with 41 S-RNases from 3 New World species Investigate long term (trans-generic) diversity at the S-RNase locus

pairwise AA divergence = 0.53 S-RNase sequence diversity Fewer trans-generic lineages S-RNase (~370 bp) for Old World S-RNases (n=4) compared to New World S- 24 OW S-RNases RNases (n=11). 41 NW S-RNases 55 Solanaceae S-RNases

ML analysis GTR + I + G

-lnL = 19783.34 Fewer trans-generic lineages (TGLs) for Old World S-RNase sequences compared to New World S-RNase sequences •! Consistent with a genetic bottleneck in the ancestor of the Old World species

Number of TGLs sensitive to both topology and sampling Re-sampling approach… (a) randomly select 24 (of 41) New World S-RNases (b) generate topology (c) count the numbers of TGLs in OW and NW Tested species of Lycium from are strongly self-incompatible based on controlled pollinations

Five species survey of the S-RNase gene…

•! Presence of ancestral S-RNase polymorphism •! Evidence of a genetic bottleneck (fewer trans-generic lineages in OW versus the NW samples) •! Re-diversification of S-RNase post bottleneck (more sites inferred to be under positive/diversifying selection in OW sample)