Genetic Analyses

The five microsatellite loci yielded between 16 and 54 alleles (mean: 32.2 ± 8.8 sd), with a total number of 152 alleles over all loci and populations. No linkage disequilibrium was observed for any pair of loci after Bonferroni correction. Therefore, further analyses were performed on multi-locus data from all five microsatellites. After Bonferroni correction, significant deviations from HWE due to a heterozygote deficit were detected for the loci LheB06 and Lhe14. Using Micro-Checker (Van Oosterhout et al. 2004), we detected a strong signal for null-alleles for these two loci, but there was no indication of errors due to stutter bands or large allele dropout for any locus. Thus, the HWE deviations at the loci LheB06 and Lhe14 could (at least partly) be due to the presence of null alleles. Values for allelic richness, observed and expected heterozygosity are given in Table 2.

The overall molecular variance among populations was 0.264 (p < 0.0001) and 0.204 were found within populations (RIS). All populations were subdivided into geographic groups following the orographic structure of the study area (for classification see Table 1). The genetic differentiation among the seven entities was strong (Rct: 0.203, Fct: 0.146, p < 0.0001 in both cases) all data are given in Table 3. A neighbour joining phenogram based on genetic distances (Cavalli-Sforza and Edvards 1967) revealed separate clusters corresponding with the respective mountain areas (Massif Central, Jura South, Jura North, Vosges, Ardennes-Eifel and Westerwald); the single populations from the Pyrenees and the Madeleine mountains are as strongly distinguished from all other samples as these clusters. Bayesian structure analysis of the data obtained for Lycaena helle individuals clearly discriminate the individuals into the same disjunct genetic groups mirroring the mountain areas. Within these groups, the degree of differentiation varies considerably. The Jura splits into two genetic groups - Southern and Northern - both of which were supported by Bayesian structure analysis with only few outliers. These two groups are also clearly discriminated by the neighbour joining phenogram (Fig. 2).

Table 2 Parameters of genetic diversity for all analysed populations of Lycaena helle in the West European study area: mean number of alleles (A), allelic richeness (AR), percentage of expected heterozygosity (He), observed heterozygostiy {HJ and private alleles (a: frequency, b: numbers)_

Area

Abbrev.

A

AR

H, [%]

H, [%]

Private allelesa[%]

Private allele sb

Pyrenees (PY)

PI

6.2

4.28

0.76

0.56

0.52

3.28

Massif Central (M)

Ml

7.4

4.28

0.71

0.52

0.40

4.60

M2

6.8

4.04

0.68

0.53

0.21

3.28

M3

7.0

4.17

0.70

0.48

0.24

2.63

M4

6.2

4.32

0.74

0.50

0.36

1.31

M5

7.2

4.86

0.79

0.57

0.20

3.95

Mean

6.92

4.33

0.73

0.53

0.38

3.2

(± s.d.)

(±0.46)

(±0.31)

(±0.04)

(±0.03)

(±0.09)

(±1.26)

Northern Jura (J)

J1

4.8

3.93

0.69

0.53

0

0

J2

7.6

4.39

0.71

0.54

0.03

1.32

J3

7.2

4.56

0.75

0.51

0.02

0.65

J4

5.8

4.56

0.71

0.61

0

0

J5

5.0

3.49

0.62

0.39

0

0

J6

6.8

4.40

0.72

0.42

0

0

Mean

6.2

4.22

0.7

0.5

0.03

0.99

(± s.d.)

(±1.17)

(±0.43)

(±0.04)

(±0.08)

(±0.01)

(±0.47)

Southern Jura (J)

J6

6.8

4.40

0.72

0.42

0

0

J7

10

5.67

0.83

0.56

0.13

3.28

J8

7.0

4.51

0.82

0.62

0.29

1.97

J9

6.0

3.87

0.68

0.52

0.59

2.63

J10

5.2

3.88

0.75

0.60

0.23

3.95

Mean

7.00

4.46

0.76

0.54

0.32

2.97

(± s.d.)

1.82

0.73

0.06

0.08

(±0.19)

(±0.84) P

Madeleine

MM

5.0

3.66

0.68

0.49

0.09

1.31 ®

Mts. (MM)

CD

Vosges (V)

VI

3.2

2.78

V2

3.6

2.80

V3

3.4

2.87

V4

3.6

2.91

Mean

3.45

2.84

(± s.d.)

(±0.19)

(±0.06)

Ardennes (A) and

Al

10.8

5.40

Eifel (E)

A2

7.0

4.35

A3

7.0

4.18

El

8.0

4.83

E2

8.2

4.82

E3

8.0

4.66

E4

6.0

4.16

Mean

7.86

4.63

(± s.d.)

(±1.51)

(±0.44)

Westerwald

W1

6.4

4.11

(WW)

W2

5.2

3.79

Mean

5.80

3.95

(± s.d.)

(±0.85)

(±0.23)

0.49

0.36

0.51

0.39

0.52

0.42

0.58

0.43

0.53

0.40

(±0.04)

(±0.03)

0.79

0.71

0.74

0.60

0.73

0.67

0.77

0.75

0.77

0.58

0.74

0.57

0.72

0.59

0.76

0.65

(±0.03)

(±0.07)

0.70372

0.55

0.675212

0.64

0.69

0.59

(±0.02)

(±0.06)

0

0

0

0

0

0

0

0

0

0

(±0.0)

(±0.0)

0.16

3.28

0.25

1.97

0.22

1.97

0.12

1.32

0.33

2.63

0.20

2.63

0.07

1.31

0.19

2.16

(±0.08)

(±0.73)

0.09

1.32

0.09

1.32

Fig. 2 Neighbour-joining phenogram based on Cavalli-Sforza and Edwards (1967) distances with bootstrap values (derived from 1,000 replicates) of 30 Lycaena helle populations from our study area. Abbreviations of localities are as in Table 1 and Fig. 1

The parameters of genetic diversity differ strongly among all classified groups. ANOVAs show significant differences among the regions for the mean number of alleles A (p = 0.014), allelic richness AR (p = 0.018), percentage of expected heterozygosity He (p = 0.016) and observed heterozygosity Ho (p = 0.007). No significant correlation was found between geographical altitude or latitude and the parameters of genetic diversity. We detected private alleles (see Material and Methods), which were found exclusively within any given region, for each mountain area except for the Vosges. In general, these alleles represent a small part of the complete gene pool. The frequencies varied between 0.02 and 0.35 within a population from one group. While populations from the Northern Jura, the Madeleine mountains and the Westerwald showed low mean frequencies of private alleles (0.03 and 0.09, respectively), the other regions showed higher frequencies (Ardennes-Eifel: 0.19, Southern Jura: 0.32, and Massif Central: 0.38). This was especially true of the Pyrenees, where private alleles exhibit about half of the total allele frequencies of the five studied microsatellite loci (0.52).

Table 3 Non-hierarchical variance analysis of Lycaena helle in five regions. All p < 0.001_

Region

st

rst

fis

ris

Ardennes-Eifel

Q.Q854

Q.Q684

Q.1342

Q.2424

Massif Central

Q.Q467

Q.1QQ9

Q.2657

Q.3Q92

Northern Jura

Q.Q553

Q.1Q65

Q.2794

Q.2975

Southern Jura

Q.1Q78

Q.2858

Q.27QQ

Q.2417

Vosges

Q.Q276

n.s.

Q.1951

Q.1195

Westerwald

Q.Q538

Q.Q761

Q.1396

n.s.

All populations

Q.1872

Q.2643

Q.17Q1

Massif Central

Ardennes

1 2 5 1G 5G 2GG Geographic distances (km)

oo d

Massif Central

5 1G 5G 2GG Geographic distances (km)

5 1G 5G 2GG Geographic distances (km)

Jura

Jura

Geographic distances (km)

Geographic distances (km)

00 d

Vosges

Vosges

00 d

5 1G 5G 2GG Geographic distances (km)

5 1G 5G 2GG Geographic distances (km)

Fig. 3 Isolation by distance in four regions studied. The ratio FST/(1 - FST) as a function of the distances between populations (logarithmic scale)

Within each regional population, all subpopulations were significantly differentiated from each other (p < 0.001). Isolation by distance was found to be significant only in the Ardennes-Eifel region (Fig. 3). No significant relationship between the number of private alleles or mean frequencies and area of potential suitable habitats (areas with Maxent values greater than 0.50 and 0.75) was detected.

Was this article helpful?

0 0

Post a comment