We screened the genetic diversity at four Northwest Iberian locations, i.e., three rivers and one reservoir (Albarellos). The most diverse locality was Tea (Table 2). Overall, most of the variation (71.39%) was due to within-population genetic diversity

M. splendens |
Lerez |
Cabe |
Tea |
Albarellos |

Lerez |
- |
*** |
*** |
NS |

Cabe |
0.48 |
- |
** |
*** |

Tea |
0.31 |
0.13 |
- |
NS |

Albarellos |
0.37 |
0.47 |
0.25 |
- |

Data in bold remained significant after sequential Bonferroni correction (P-values estimated after 10,000 permutations). Above diagonal: combined probabilities for each pairwise population differentiation (exact differentiation test). * P > 0.05, ** P < 0.01, *** P < 0.001 NS non-significant

Data in bold remained significant after sequential Bonferroni correction (P-values estimated after 10,000 permutations). Above diagonal: combined probabilities for each pairwise population differentiation (exact differentiation test). * P > 0.05, ** P < 0.01, *** P < 0.001 NS non-significant

O. curtisii |
Lerez |
Cabe |
Tea |
Deza |
Arnoia |

Lerez |
- |
*** |
*** |
*** |
*** |

Cabe |
0.21 |
- |
*** |
*** |
*** |

Tea |
0.15 |
0.40 |
- |
*** |
*** |

Deza |
0.12 |
0.31 |
0.24 |
- |
*** |

Arnoia |
0.16 |
0.21 |
0.34 |
0.26 |
- |

Data in bold remained significant after sequential Bonferroni correction (P-values estimated after 10,000 permutations). Above diagonal: combined probabilities for each pairwise population differentiation (exact differentiation test). * P > 0.05, ** P< 0.01, *** P < 0.001 NS Non-significant

Data in bold remained significant after sequential Bonferroni correction (P-values estimated after 10,000 permutations). Above diagonal: combined probabilities for each pairwise population differentiation (exact differentiation test). * P > 0.05, ** P< 0.01, *** P < 0.001 NS Non-significant

(Table 5). There was significant population structuring according to sampling location, as suggested by both the test of exact differentiation (c2=213.77, d.f. = 50, P < 0.001) and F-statistics (6 = 0.29, CI=(0.22-0.35)). Regarding population-wise comparisons, the extent of differentiation ranged from moderate (Tea-Cabe, FST=0.13) to strong (Lerez-Cabe, FST = 0.48). Albarellos was the only locality not significantly differentiated from any other (Table 3). No grouping was significantly differentiated from AMOVA analysis (Table 5), although river Cabe clearly separated from the rest of samples, as revealed by the UPGMA tree (Fig. 2a). Geographic and genetic distances were not correlated (Mantel test, r=0.38, P=0.26). Results from no-admixture clustering with STRUCTURE 2.1 indicated that the most likely grouping of Macromia individuals was K = 2. The first two sampled localities (Lerez and Albarellos) appeared as purely belonging to a first population (POP1) (white, Fig. 3). The third sampled site (Tea) contained (1) 50% of individuals directly assigned to a second population, POP2 (grey, Fig. 3), (2) 11% of individuals directly assigned to POP1, and (3) 39% of individuals whose genetic pattern was intermediate between POP1 and POP2, but with a marked trend towards POP2. The fourth sampled locality (Cabe) was almost totally composed of individuals belonging to POP2.

Table 5 Analysis of molecular variance (AMOVA) across a range of putative population groupings

Percentage of variation accounted for (F)

Table 5 Analysis of molecular variance (AMOVA) across a range of putative population groupings

Percentage of variation accounted for (F)

Among | |||||

populations |
Within- | ||||

Grouping |
Population |
Among |
within groups |
populations | |

Species |
criterion |
grouping |
groups (FCT) |
Fsc) |
(Fst) |

Macromia |
- |
All in one |
- |
28.61 (0.29)*** |
71.39 |

splendens |
group | ||||

River basin |
[Cabe/Tea/ Albarellos] [Deza] [Lérez] |
16.01 (0.16) |
19.61 (0.23)*** |
64.38 (0.36)*** | |

Reservoir vs. |
[Albarellos] |
8.64 (0.08) |
24.45 (0.27)*** |
66.91 (0.33)*** | |

river |
[all remaining POPs] | ||||

STUCTURE |
[Albarellos/ |
21.87 (0.22) |
13.51 (0.17)*** |
64.62 (0.35)*** | |

results |
Lérez] [Tea/Cabe] | ||||

UPGMA |
[Cabe] [all |
-4.69 (-0.05) |
31.96 (0.30)*** |
72.73 (0.27)*** | |

tree |
remaining POPs] | ||||

Oxygastra |
- |
All in one |
- |
22.17 (0.22)*** |
77.83 |

curtisii |
group | ||||

River basin |
[Cabe/Tea/ Arnoia] [Deza] [Lérez] |
-15.05 (-0.15) |
34.9 (0.30)*** |
80.17 (0.19)*** | |

UPGMA |
[Cabe/Arnoia] |
11.02 (0.11)* |
15.13 (0.17)*** |
73.86 (0.26)*** | |

tree |
[Tea/Deza/ Lérez] | ||||

UPGMA |
[Cabe/Arnoia] |
10.09 (0.10)** |
13.56 (0.15)*** |
76.35 (0.24)*** | |

Lérez] |

* P=0.1, ** P=0.061, *** P < 0.001. Values in italics correspond to FST

* P=0.1, ** P=0.061, *** P < 0.001. Values in italics correspond to FST

We surveyed the genetic diversity of this species in five Northwest Iberian riverine localities. The most diverse locality was Lerez (Table 1). The within-population genetic diversity accounted for most of the variation (77.83%). There was significant population differentiation between sampling localities, as suggested by both the test of exact differentiation (c2=446.69, d.f. = 66, P < 0.001) and F-statistics (0 = 0.16, CI=(0.12-0.20)). Regarding population-wise comparisons, the extent of differentiation ranged from moderate (Lerez-Deza, FST=0.12) to strong (Tea-Cabe, F = 0.40) (Table 4). Geographic distance and genetic divergence were significantly

0.15

0.10

0.05

- ALBARELLOS _ LEREZ

0.00

b 100

LEREZ

DEZA TEA CABE ARNOIA

0.08

0.06

0.04

0.02

0.00

Fig. 2 UPGMA dendrogram based on Nei's unbiased genetic distance. Numbers at nodes are bootstrap values < 50% (10,000 pseudoreplicates). (a) Macromia splendens, (b) Oxygastra curtisii a

1(1) 3(1) 5(1) 7(2) 9(2) 11(3) 13(3) 15(3) 17(3) 19(3) 21(3) 23(3) 25(3) 27(3) 29(3) 31(4) 33(1) 35(4) 37(4) 39(4; 2(1) 4(1) 5(1) 3(2) 10(2) 12(3) 14(3) 16(3) 13(3) 20(3) 22(3) 24(3) 26(3) 26(3) 30(4) 32(d) 34(4) 36(4) 33(4)

Fig. 3 Results of STRUCTURE no-admixture clustering for Macromia splendens. Each individual is represented by a vertical column and the proportion of each of the two colors indicated the putative population they derive from. The most likely number of population (K=2) was determined following Evanno et al. (2005). Numbers in the X axis identify each analyzed specimen (n = 38), giving its geographical origin in parentheses: (1) Lerez, (2) Albarellos, (3) Tea, and (4) Cabe

Fig. 2 UPGMA dendrogram based on Nei's unbiased genetic distance. Numbers at nodes are bootstrap values < 50% (10,000 pseudoreplicates). (a) Macromia splendens, (b) Oxygastra curtisii

1(1) 3(1) 5(1) 7(2) 9(2) 11(3) 13(3) 15(3) 17(3) 19(3) 21(3) 23(3) 25(3) 27(3) 29(3) 31(4) 33(1) 35(4) 37(4) 39(4; 2(1) 4(1) 5(1) 3(2) 10(2) 12(3) 14(3) 16(3) 13(3) 20(3) 22(3) 24(3) 26(3) 26(3) 30(4) 32(d) 34(4) 36(4) 33(4)

correlated in the area studied (Mantel test, r=0.61, P < 0.05). The phylogenetic reconstruction showed two differentiated clades (Fig. 2b), in agreement with the geographic location.

Was this article helpful?

## Post a comment