Landscape Genetics

1
Landscape Genetics
2
What is landscape genetics?

• How do landscape features affect genetic
  structure of populations?
• Analyze spatial genetic data
• Combines tools of population genetics and
  landscape ecology
• Does not require a priori designation of
  populations

3
Common spatial patterns

• Isolation by distance
• Clines
• Discontinuities (boundaries)
• Metapopulations
• Random

4
Genetic tools

• Individual is sampling unit
• Highly polymorphic biparentally inherited markers
• Microsats
• SNPs
• AFLPs
• Allozymes

5
Spatial analysis tools

• Mantel test/regression
• Tests correlation between 2 variables in matrices
• Partial Mantel test
• Tests correlation between genetic relatedness
  matrix and more than one matrix of environmental
  variables
• Bayesian/maximum-likelihood clustering
• Assigns individuals to populations
• Spatial autocorrelation
• Is genotype of individual dependent on genotypes
  of neighboring individuals?
• Can be positive or negative from -1.0 to 1.0)
• Morans I (spatial covariance/total variance)
• Wombling
• Detects boundaries through analysis of allele
  frequencies
• Monmoniers algorithm
• Detects boundaries through analysis of genetic
  distances
• Spatial Analysis of Variance (SAMOVA)
• Defines populations that are geographically
  homogeneous and maximally differentiated from
  each other

6
(No Transcript)
7
Wombling
Monmoniers algorithm
8
Delaunay triangulation
Voronoi tesselation
9
Delaunay triangulation
10
Connectivity/dispersal

• Spatial analysis of salamander gene flow
• Spear et al. Molecular Ecology 142553-2564
  (2005)
• Tested whether landscape variables were
  correlated with population structure (FST)

11
Straight-lineroute
Wetland likelihoodroute
12
Combination least slope/wetland likelihood route
Stepping-stone route
Least slope route
13
(No Transcript)
14
Wombling human allele frequencies in Europe
15
(No Transcript)
16
Friction maps

• Assumes a cost or resistance to travel across
  a landscape
• Grid cells on a map can be assigned a cost for
  one or more variables
• Least-cost path can be determined
• Connectivity can be assessed

17
Broquet et al. Molecular Ecology 21877-889
(2006)
18
Euclidean
Least cost
Unlogged
Logged
19
Roe deer in a fragmented forest landscape(Coulon
et al. 2004 Mol. Ecol. 13284-2850)
20
(No Transcript)
21
Euclidean
Least cost
Females
Males
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)