Table 2

Results of outlier detection among the 83 AFLP markers in larvae of the large pine weevil using the frequentist method Dfdist and the Bayesian inference method BayeScan.

Method of detection

Frequentist

Bayesian inference


Dataset

Locus

p-value

FST

Posterior probability

A

FST


Geography (Structure 11)

523

0.000

0.231

1

2.010

0.221

68

0.000

0.338

1

1.790

0.191

38

0.000

0.276

1

1.810

0.194

10

0.000

0.248

0.758

1.000

0.105

63

0.004

0.180

0.999

1.520

0.156

13

0.099

0.061

0.974

1.320

0.136

47

0.045

0.080

0.931

1.190

0.122


Geography + host-plant

38

0.000

0.259

1

2.090

0.208

(Structure 21)

52

0.000

0.256

1

2.180

0.220

63

0.000

0.225

1

1.950

0.186

68

0.000

0.231

0.999

1.650

0.151

10

0.000

0.181

0.743

0.893

0.082

30

0.016

0.088

0.908

1.110

0.098

33

0.018

0.095

0.876

1.030

0.091

27

0.102

0.066

0.882

1.040

0.092

13

0.069

0.054

0.981

1.300

0.116

47

0.043

0.065

0.865

1.030

0.092


Local host-plant differentiation

Regions

Finland2 (Structure 31)

Limousin (Structure 41)

27

0.000

0.217

0.915

1.460

0.222

Ardeche2 (Structure 51)


1As in Figure 1.

2No outliers were found with the significance level used.

3Bold type indicates markers that are detected by both methods with a type-I error (α) = 0.0006 for Dfdist, and with a posterior probability > 0.79 for BayeScan (see text for explanation about these values).

Manel et al. BMC Evolutionary Biology 2009 9:288   doi:10.1186/1471-2148-9-288

Open Data