Speciation in the ocean could differ from terrestrial environments due to fewer barriers to gene flow. Hence, sympatric speciation might be common, with American and European eel being candidates for exemplifying this. They show disjunct continental distributions on both sides of the Atlantic, but spawn in overlapping regions of the Sargasso Sea from where juveniles are advected to North American, European and North African coasts. Hybridization and introgression is known to occur, with hybrids almost exclusively observed in Iceland. Different speciation scenarios have been suggested, involving either vicariance or sympatric ecological speciation. Using RAD sequencing and whole-genome sequencing data from parental species and F1 hybrids, we analyzed speciation history based on the Joint Allele Frequency Spectrum (JAFS) and PSMC (pairwise sequentially Markovian coalescent) plot. JAFS supported a model involving a split without gene flow 150,000 – 160,000 generations ago, followed by secondary contact 87,000 – 92,000 generations ago, with 64% of the genome experiencing restricted gene flow. This supports vicariance rather than sympatric speciation, likely associated with Pleistocene Glaciation cycles and ocean current changes. Whole genome PSMC analysis of F1 hybrids from Iceland suggested divergence 200,000 generations ago and indicated subsequent gene flow rather than strict isolation. Finally, simulations showed that results from both approaches (JAFS and PSMC) were congruent. Hence, there is strong evidence against sympatric speciation in North Atlantic eels. These results reiterate the need for careful consideration of cases of possible sympatric speciation, as even in seemingly barrier-free oceanic environments palaeoceanographic factors may have promoted vicariance and allopatric speciation.

The data encompasses a Genepop (Atlantic_eels_Genepop.txt) and a VCF file (Atlantic_eels.vcf) consisting of 9,481 SNPs in European and American eel filtered from a previously published dataset generated by RAD sequencing (Jacobsen MW, Pujolar JM, Bernatchez L, et al. (2014) Genomic footprints of speciation in Atlantic eels (Anguilla anguilla and A. rostrata). Molecular Ecology 23, 4785–4798). 

In addition, commands (*.sh files) are included for simulations of pure and hybrid eel genomes using ms (Hudson RR (2002) Generating samples under a Wright-Fisher neutral model of genetic variation. Bioinformatics 18, 337-338).

In the SNP dataset (Genepop and VCF files) the following individuals are European eel: LEG11 LEG9 RHG12 RHG13 RHG14 RHG16 RHG17 RHG3 RHG5 RHG6 RHG8 LEG12 RHG9 VG11 VG12 VG13 VG14 VG15 VG16 VG17 VG18 VG19 LEG13 VG20 LEG14 LEG15 LEG16 LEG20 LEG7 LEG8.

The following individuals are American eels: NEG19 RBA14 RBA15 RBA16 RBA17 RBA19 RBA1 RBA21 RBA22 RBA2 RBA4 NEG20 RBA5 RBA6 RBA7 RBG11 RBG20 RBG27 RBG28 STJ33 STJ34 NEG24 NEG25 NEG29 NEG30 NEG4 RBA12 RBA13.

For the simulation commands IM denotes "Isolation with migration", SI denotes "Strict isolation", and SCG denotes "Secondary contact with exponential growth". AM denotes American eel, EU denotes European eel and F1 denotes F1 hybrids between the two species.