Phylogeography and phenotypic wing shape variation in a damselfly across populations in Europe.

BACKGROUND: Describing geographical variation in morphology of organisms in combination with data on genetic differentiation and biogeography can provide important information on how natural selection shapes such variation. Here we study genetic structure using ddRAD seq and wing shape variation using geometric morphometrics in 14 populations of the damselfly Lestes sponsa along its latitudinal range in Europe. RESULTS: The genetic analysis showed a significant, yet relatively weak population structure with high genetic heterozygosity and low inbreeding coefficients, indicating that neutral processes contributed very little to the observed wing shape differences. The genetic analysis also showed that some regions of the genome (about 10%) are putatively shaped by selection. The phylogenetic analysis showed that the Spanish and French populations were the ancestral ones with northern Swedish and Finnish populations being the most derived ones. We found that wing shape differed significantly among populations and showed a significant quadratic (but weak) relationship with latitude. This latitudinal relationship was largely attributed to allometric effects of wing size, but non-allometric variation also explained a portion of this relationship. However, wing shape showed no phylogenetic signal suggesting that lineage-specific variation did not contribute to the variation along the latitudinal gradient. In contrast, wing size, which is correlated with body size in L. sponsa, had a strong negative correlation with latitude. CONCLUSION: Our results suggest a relatively weak population structure among the sampled populations across Europe, but a clear differentiation between south and north populations. The observed geographic phenotypic variation in wing shape may have been affected by different local selection pressures or environmental effects.

The genomic response to urbanization in the damselfly Ischnura elegans.

The complex and rapid environmental changes brought about by urbanization pose significant challenges to organisms. The multifaceted effects of urbanization often make it difficult to define and pinpoint the very nature of adaptive urban phenotypes. In such situations, scanning genomes for regions differentiated between urban and non-urban populations may be an attractive approach. Here, we investigated the genomic signatures of adaptation to urbanization in the damselfly Ischnura elegans sampled from 31 rural and urban localities in three geographic regions: southern and northern Poland, and southern Sweden. Genome-wide variation was assessed using more than 370,000 single nucleotide polymorphisms (SNPs) genotyped by ddRADseq. Associations between SNPs and the level of urbanization were tested using two genetic environment association methods: Latent Factors Mixed Models and BayPass. While we found numerous candidate SNPs and a highly significant overlap between candidates identified by the two methods within the geographic regions, there was a distinctive lack of repeatability between the geographic regions both at the level of individual SNPs and of genomic regions. However, we found "synapse organization" at the top of the functional categories enriched among the genes located in the proximity of the candidate urbanization SNPs. Interestingly, the overall significance of "synapse organization" was built up by the accretion of different genes associated with candidate SNPs in different geographic regions. This finding is consistent with the highly polygenic nature of adaptation, where the response may be achieved through a subtle adjustment of allele frequencies in different genes that contribute to adaptive phenotypes. Taken together, our results point to a polygenic adaptive response in the nervous system, specifically implicating genes involved in synapse organization, which mirrors the findings from several genomic and behavioral studies of adaptation to urbanization in other taxa.