Hybridization and introgression have played important roles in the history of various species, including lineage diversification and the evolution of adaptive traits. Hybridization can accelerate the development of reproductive isolation between diverging species, and thus valuable insight into the evolution of reproductive barrier formation may be gained by studying secondary contact zones. Hedgehogs of the genus Erinaceus, which are insectivores sensitive to changes in climate, are a pioneer model in Pleistocene phylogeography. The present study provides the first genome-wide SNP data regarding the Erinaceus hedgehogs species complex, offering a unique comparison of two secondary contact zones between Erinaceus europaeus and E. roumanicus. Results confirmed diversification of the genus during the Pleistocene period and detected a new refugial lineage of E. roumanicus outside the Mediterranean region, most likely in the Ponto-Caspian region. In the Central European zone, the level of hybridization was low, whereas in the Russian-Baltic zone, both species hybridise extensively. Asymmetrical gene flow from E. europaeus to E. roumanicus suggests that reproductive isolation varies according to the direction of the crosses in the hybrid zones. However, no loci with significantly different patterns of introgression were detected. Markedly different pre- and post-zygotic barriers, and thus diverse modes of species boundary maintenance in the two contact zones, likely exist. This pattern is probably a consequence of the different ages and thus of the different stages of evolution of reproductive isolating mechanisms in each hybrid zone.

The sample list includes 33 specimens of Erinaceus europaeus, 48 of E. roumanicus, five of E. concolor, two E. amurensis and one Hemiechinus auritus and Atelerix albiventris. All tissue samples were obtained from roadkill animals or from animals that died in rescue centres. No animals were sacrificed for the purpose of this study. The samples from Russia were imported to the Czech Republic under requested veterinary conditions for import with reference numbers SVS/2019/061603-G. The remainder of the samples were collected before 2013. The tissues were fixed in 96% ethanol and stored at -20 °C in the laboratory which has approval (No CZ 11712934) for storage and usage of animal material according to § 48(1)(i) of Act No 166/1999s concerning veterinary care and amending certain related laws, as amended, pursuant to Article 17(1) of Regulation of the European Parliament and the Council (EC) No 169/2009 and Article 27(1) of Commission Regulation (EU) No 142/2011. The sample names, their geographic coordinates and species determination are listed in the metadat.txt file.

DNA was extracted using spin-column protocol (Qiagen DNeasy Blood and Tissue kit, Qiagen). Mitochondrial control region sequences and combinations of microsatellite loci were used for species determination. Restriction‐site associated DNA (RAD) libraries were prepared for 50 samples using the restriction enzyme Sbfl1 and performed at the European Molecular Biology Laboratory in Heidelberg. The codewords used in multiplexing are specified in the metadata.txt file. For DNA fragmentation, adapter ligation, purification and fragment size segregation, the protocol of the NEBNext® Ultra™ DNA Library Prep Kit for Illumina® (Illumina Inc.) was used. Two prepared paired-end libraries were sequenced at the European Molecular Biology Laboratory (Heidelberg) using the Illumina HiSeq2000 sequencer, in two lanes separately, according to the manufacturer's protocol (Illumina Inc.). A single-end library was subsequently prepared for 40 samples from the Russian-Baltic contact zone, with the same restriction enzyme. The library was then sequenced using Illumina NextSeq at the European Molecular Biology Laboratory (Heidelberg). The sequencing method applied to each sample is specified in the metadata.txt file.

Sequence demultiplexing based on individual specific codewords was performed using Sabre (https://github.com/najoshi/sabre), and read quality assessment was performed in FastQC 0.11.5 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). The three sequencing batches were processed independently, producing three complementary sets of fastq files: fastq2016.tar and fastq2018.tar contain the paired-end sequences from the first 50 samples, sequenced twice; and fastq2020.tar contains the single-read sequences from the 40 additional samples sequenced.

The data set is composed of several files:

metadata.txt describes the samples.

fastq2016_X.tar (with "X" between 1 and 5) extract a directory named fastq2016 with the compressed fastq files corresponding to the first sequencing batch of the 50 first samples. Note that there are two files per sample, for the forward and the reverse reads.

fastq2018_X.tar (with "X" between 1 and 3) extract a directory named fastq2018 with the compressed fastq files corresponding to the second sequencing batch of the same 50 first samples. Note that the file names in this directory are the same as in fastq2016, because they are the same samples.

fastq2020_X.tar (with "X" between 1 and 6) extract a directory named fastq2020 with the compressed fastq files corresponding to the sequencing of 40 additional samples. Here there is one file per sample, because they were sequenced with single reads.