Active transposable elements (TEs) may result in divergent genomic insertion and abundance patterns among conspecific populations. Upon secondary contact, such divergent genetic backgrounds can theoretically give rise to classical Dobzhansky-Muller incompatibilities (DMI), thus contributing to the evolution of endogenous genetic barriers and eventually cause population divergence. We investigated differential TE abundance among conspecific populations of the non-biting midge Chironomus riparius and evaluated their potential role in causing endogenous genetic incompatibilities between these populations. We focussed on a Chironomus-specific TE, the minisatellite-like Cla-element, whose activity is associated with speciation in the genus. Using a newly generated and annotated draft genome for a genomic study with five natural C. riparius populations, we found highly population-specific TE insertion patterns with many private insertions. A significant correlation of the pairwise FST estimated from genome-wide single nucleotide polymorphisms (SNPs) and the FST estimated from TEs, is consistent with drift as the major force driving TE population differentiation. However, the significantly higher Cla-element FST level due to a high proportion of differentially fixed Cla-element insertions also indicates selection against segregating (i.e. heterozygous) insertions. With reciprocal crossing experiments and fluorescent in-situ hybridisation of Cla-elements to polytene chromosomes, we documented phenotypic effects on female fertility and chromosomal mispairings. We propose that the inferred negative selection on heterozygous Cla-element insertions may cause endogenous genetic barriers and therefore acts as DMI among C. riparius populations. The intrinsic genomic turnover exerted by TEs may thus have a direct impact on population divergence that is operationally different from drift and local adaptation.

find-uniq-BLAST-hits.plThis script filters duplicates from a BLAST table (-outfmt 6). When the identical data (in our case Pool-Seq reads) is blasted against similar reference sequences (in our case variants of a specific transposable element,i.e. the Cla-element), identical query sequences might fit to multiple reference sequences. This produces duplicates in the merged BLAST output. With this script retains the first occurrence of a query sequence and outputs a table of the same format as the input (BLAST -outfmt 6).PosinExons.pyAnnotation of genomic locations that fall into exons.PosinGenes.pyAnnotation of genomic locations that fall into genes (including introns).MF_50Mio_pe.sorted.bam50 Mio random pe reads of Pool-Seq data MF (Rhone-Alpes) mapped to Crip Laufer draft genomeMG_50Mio_pe.sorted.bam50 Mio random pe reads of Pool-Seq data MG (Hessen, Germany) mapped to Crip_Laufer draft genome. Raw reads available at EMBL-EBI ENA ERS1590059NMF_50Mio_pe.sorted.bam50 Mio random pe reads of Pool-Seq data NMF (Lorraine) mapped to Crip_Laufer draft genome. Raw reads available at EMBL-EBI ENA ERS1590060SI_50Mio_pe.sorted.bam50 Mio random pe reads of Pool-Seq data SI (Piemont, Italy) mapped to Crip_Laufer draft genome. Raw reads available at EMBL-EBI ENA ERS1590061SS_50Mio_pe.sorted.bam50 Mio random pe reads of Pool-Seq data MG (Andalusia, Spain) mapped to Crip_Laufer draft genome. Raw reads available at EMBL-EBI ENA ERS1590062