Ecological Genetics

1
Ecological Genetics

• Part 1

2
Background

• Turesson (1922-1930) common garden experiments
  of Swedish plants
• Work extended by Clausen, Keck Heisey 1940s
  plants from an altitudinal gradient in a common
  garden
• 1950s animals (Cain Sheppard) banded snails
• Ecological Genetics Group in 1956 in UK
• First used in print by Ford (1964) Ecological
  Genetics
• Past 25 years increasing interest in ecological
  genetics advances in molecular marker
  technologies

3
What is Ecological Genetics?

• Interface between ecology, evolution and
  genetics (Conner Hartl 2004)
• Interaction of large-scale geographic patterns
  of demography with genetic dynamics among small,
  partially isolated, and potentially locally
  adapted populations. (Lowe et al 2004)

4
What can Ecological Genetics tell us?

• Adaptive radiation evolution of ecological
  diversity within rapidly multiplying phylogenetic
  lineages
• Alien invasions
• Tracking the release of GM organisms
• Interactions between wild and domesticated
  species
• Migration and gene flow between once continuous
  populations
• Local adaptation
• Environmental perturbations what can glaciation
  events tell us about climate change?

5
Why are these studies important?

• Conservation
• Rare and threatened species
• Common species increasingly important
• Management
• Agricultural/forestry landscapes

6
Markers

• Vast array of molecular markers now available
• Lees lecture this afternoon
• Good science depends on
• Interesting/important question most appropriate
  molecular marker
• Not all markers are created equal
• Codominant vs. dominant
• However, most appropriate marker may not be
  available (e.g. SSRs)
• Use another type of marker
• Refine your question

7
The Ideal Ecological Genetic Marker

• Detect qualitative or quantitative variation
• Present/absence
• Discrete variation e.g. Low vs. high variation
• No environmental or development influences
• 3 environments same genotype
• Juvenile adult
• Simple codominant inheritance
• Both alleles visible in heterozygous individuals
• Beware of null alleles (alleles that are not
  expressed)

(Weising et al. 1995)
8
The Ideal Ecological Genetic Marker

• Detect silent nucleotide changes
• Changes in coding regions that result in
  synonymous amino acid substitutions into protein
  sequences
• e.g. GTT GTC GTA GTG valine
• Detect changes in coding and non-coding regions
  of genome
• Randomly distributed markers
• Detect evolutionary homologous changes
• Similar due to common ancestor

(Weising et al. 1995)
9
Sampling
Organisms are spatially and temporally
distributed
10
Sampling
Variation occurs within and among
populations
11
Sampling

• Often driven by
• Objective of your study
• Research funds and facilities available
• Sampling strategy should aim to maximise
  efficiency
• Provide best statistical estimates of parameters
  under study at lowest possible cost

12
Sampling

• What is a population?
• The eternal question.
• Thomas lecture yesterday
• Three definitions in ecological genetics
• Statistical universe of organisms under study
• Ecological group of organisms of one taxon in
  particular area at particular time (ie.
  biological population or provenance)
• Genetical individuals that are connected by
  gene flow (ie. gene pool)

13
Sampling
Genetic, ecological and statistical, populations
coincide
Samples statistical population, but ecological
and genetic populations do not coincide
Ecological and genetic populations coincide, but
not statistical population
Sample
Statistical population
(Lowe et al. 2004)
14
Simple Random Sampling

• One of the most straightforward probability
  sampling techniques
• Statistical population consisting of N sampling
  units from which n units selected with every unit
  having same chance of being chosen

15
Simple Random Sampling
(Lowe et al. 2004)
16
Simple Random Sampling

• In practice, often difficult to follow this
  approach
• Need to label all trees
• Know ecological distribution all populations
• Use different sampling approach
• Accessibility sampling

17
Accessibility Sampling
(Lowe et al. 2004)
18
Simple Random Sampling

• Use different sampling approach
• Accessibility sampling
• Haphazard sampling - opportunistic
• Judgemental sampling - collectors experience

19
Stratified Random Sampling

• Powerful sampling technique
• Statistical sampling population of N units is
  divided into L non-overlapping strata, which
  together comprise whole population

20
Stratified Random Sampling
(Lowe et al. 2004)
21
Stratified Random Sampling

• If strata are well defined, improves precision of
  estimates for mean and C.I. for whole of
  population
• But, may get differences in sampling rates in
  each stratum
• Organism availability
• Collector differences
• Are boundaries real or artificial?
• Need to make decisions regarding
• Number of strata
• Allocation of sampling units to strata
• Sampling costs vs. number of samples required

22
Systematic Sampling

• Common and convenient
• Sample at fixed points on a line, grid or
  physical feature (e.g. road, river)
• Justification
• Simplicity for use in field
• Desire to sample evenly across a region
• Minimize sampling of closely related organisms
• Particularly useful if suspect a gradient or
  cline, e.g. hybrid zone

23
Systematic Sampling

• Considerations
• Periodic variation can lead to bias in estimating
  mean and variation of a population
• In practice, organisms, particularly plants, tend
  to be clumped and irregular

24
Sampling Practicalities

• Sampling design
• Ecology
• Distribution
• Reproductive biology
• Have other researchers used similar species
• How to collect organism?
• Tall canopy trees
• What material to collect?
• Leaf, seed, pollen depends problem molecular
  marker

25
Sampling Practicalities

• How to transport material to lab?
• Legal and ethical collections
• Voucher specimens
• Location data GPS
• Field safety

26
Common Population Genetic Measures

• Allele frequencies
• Allele number
• Gene diversity
• Genetic distance
• Covered by Thomas and Kevin

27
Within-population Sampling Allele Frequencies

• Allele frequencies commonly calculated in
  ecological genetic studies
• To estimate sampling size, need to know allele
  frequencies
• BIG PROBLEM often we dont know the allele
  frequencies, we are going to determine them as
  part of the study

28
Within-population Sampling Allele Frequencies

• Estimate on worse case scenario for codominant,
  diallelic locus
• p frequency of allele, say 0.5
• 95 C.I. requires sample size of

400 independent gametes (i.e. 200 diploid
organisms)


29
Within-population Sampling Allele Frequencies

• If accept a higher error margin, or if allele
  frequency varies from 0.5 sample size required
  changes

From Lowe et al. 2004
30
Within-population Sampling Allele Frequencies

• Dont always have codominant markers available
• Many studies use dominant markers such as AFLPS
  need more samples

From Lowe et al. 2004
31
Within-population Sampling Allele Number

• Important criterion for conservation management,
  particularly
• Germplasm collections
• Priority setting
• If population has two alleles (A1, A2) at
  frequencies p1 and p2, probability that a random
  sample of n gametes contains at least one copy of
  each allele is
• PA1, A2 1(1p1)n (1p2)n (1p1p2)n

32
Within-population Sampling Allele Number

• If p10.95 and p20.05 then 59 gametes are
  required to obtain one copy of each allele with
  95 certainty
• If frequency of commonest allele rises to 0.99,
  then need to increase sample to 300 gametes
• Frequency of rarest allele is important for this
  measure
• Number of alleles per locus also important
• E.g. 20 alleles per locus each with frequency of
  0.05 requires random sample of 120 gametes to
  provide 95 certainty of one copy of each allele
  being sampled

33
Within-population Sampling Gene Diversity

• Widely used measure of genetic variation
• Need optimal sampling of both loci and
  individuals
• Nei (1978) recommends large numbers (gt50
  codominant loci)
• Refined estimate in 1987 to 25 loci 20-30
  individuals per locus
• Marker choice can have an important influence

34
Within-population Sampling Genetic Distance

• When gene diversities are gt0.1 require large
  numbers of individuals (gt50) to construct robust
  dendograms
• Although recommended use gt50 loci, often only
  15-30 codominant loci are used
• Often hard to find suitable markers
• Sampling costs often a consideration
• Often 20-30 individuals per locus sampled

35
Within-population Sampling Gene Flow

• Parentage analysis
• Identify parents (usually fathers) and patterns
  of gene movement
• Need to know
• Exclusion probability of marker
• Level of gene movement wish to detect
• Number of potential fathers within study
  population
• If maternal genotype known, exclusion probability
  of 0.8, 5 of gametes to be estimated and 5
  potential fathers THEN 300 progeny sampled to
  ensure 95 C.I. If marker probability increases
  to 0.9, then need 200 progeny

36
Within-population Sampling Gene Flow

• For each mother sample more progeny than
  potential fathers
• Best markers have allele frequencies in almost
  equal proportions, i.e. rare alleles not very
  useful

37
Among-population Sampling

• Depends on question being asked
• Pattern of genetic variation expected
• Similar species studies
• Biological traits of species being studied
• Rarely possible sample all populations for common
  species
• Accessibility
• Financial/time constraints

38
Among-population Sampling

• Better to have more populations with fewer
  samples OR fewer populations with more samples?
• For genetic diversity better to sample as
  widely as possible across geographic and
  ecological range

39
More to come in Part II

• Questions