Saimaan harjuskantojen perinnlliset erot ja istutusten aikaansaamat muutokset

1
The relative role of drift and selection in
life-history evolution a case study from
recently founded populations of grayling
Mikko Koskinen, Thrond Haugen Craig Primmer
2
Outline

• Background to Darwinian vs neutral
  evolutionary theories
• Presentation of the framework of this study
• Results and conclusions

3
Natural Selection vs Random Drift?

• Transplantation experiments suggest that
  selection is an efficient evolutionary force

- Anolis lizards were introduced onto small
islands - Populations differentiated from each
other over 10-14 years according to the recipient
islands vegetation (Losos et al. 1997, Nature)
4
Natural Selection vs Random Drift?

• Comparisons of phenotypic differentiation and
  selectively neutral differentiation (e.g. from
  non-coding DNA microsatellites) suggest that
  selection is an efficient evolutionary force

1
- Mean quantitative genetic (QST) differentiation
generally exceeded neutral marker gene (FST)
differentiation (Merilä Crnokrak 2001, J.
Evol. Biol.)
QST
0.1
0.01
0.01
0.1
1
FST
5
Natural Selection vs Random Drift?
S. Wright

• In stark contrast with the Darwinian view,
  influential theories (e.g. Wright 1931, Kimura
  1995) suggest that drift is the dominant
  evolutionary force in finite populations
• It is fair to say that the neutral drift
  hypothesis has been among the most controversial
  issues in evolutionary biology in the last 50
  years, and is empirically understudied

6
Outline

• Background to Darwinian vs neutral
  evolutionary theories
• Presentation of the framework of this study
• Results and conclusions

7
Study system
European grayling, Thymallus thymallus
Norway
8
The plan

• To measure quantitative genetic differences (QST)
  between the populations using common-garden
  experiments (six early life-history traits)
• Three temperatures, three populations with
  half-sib design
• Four unique females mated with each male (28
  families per population)
• (Spitze et al 1993, Genetics)
• Variance components from mixed-model Anova
• 95 CI from non-parametric bootstrapping
• To measure neutral genetic differences (FST),
  i.e. the effect of drift, using 17 microsatellite
  DNA loci

9
The plan

• To investigate the demographic history of the
  populations using microsatellites, and to use
  that for interpreting how the results relate to
  the Darwinian vs neutral evolutionary
  theories
• To test the null-hypothesis of neutral evolution
  of the six traits using
• (Lande 1976, Evolution)
• -Ne effective population size
  (maximum-likelihood estimate from microsatellite
  data)
• -s2GB additive genetic variance between
  populations
• -s2GW additive genetic variance within
  populations (among sire var comp)
• -h2 narrow-sense heritability in a given
  population and environment
• -t divergence time of populations

10
Outline

• Background to Darwinian vs neutral
  evolutionary theories
• Presentation of the framework of this study
• Results and conclusions

11
Results - neutrality tests

• Neutral evolution was rejected for the majority
  of the trait
• Recall

12
Results - neutrality tests

• Extremely low Ne estimates, not compatible with
  sexual reproduction, would have been required for
  drift to dominate over selection F1, ?
  3.84 for P ? 0.05

13
Results - QST vs FST

• Population differences based on quantitative
  traits (QST) often strikingly exceeded the
  analogous measures based on microsatellites (FST)

Les vs Ht
FST
length at termination
yolk-sac volume
growth rate
incubation time
swim-up length
hatching length
0.0
1.0
0.5
14
Results - demographic history

• Effective sizes of the populations were small
• Microsatellite diversity within populations was
  low
• The populations have historically experienced
  severe bottlenecks

Show examples of Ne sampling distributions and
likewise for N0/N1
15
Conclusions

• The evolution of the phenotypic differences
  between the populations was dominantly due to
  natural selection -neutrality tests F-test and
  Ne(sign) estimates -QST vs FST comparisons
• Provide Fst/QstFst

16
Conclusions

• However, also drift had a notable effect (FST
  0.05-0.21)
• The dominating effect of selection is interesting
  in the light of the demographic history of the
  populations.
• According to the influential neutral theory,
  the low Nes and bottlenecks should have
  emphasized the effect of drift

17
Acknowledgements

• Thanks to Juha Merilä, Mark Beaumont, Asbjørn
  Vøllestad, Peter Crnokrak, Andrew Hendry, Martin
  Lascoux and Nick Smith for helpful comments!