At Saint Crépin location (French Alps), where sympatry between the tetraploid Juniperus thurifera and the diploid Juniperus sabina occurs, three individuals with an atypical morphology, have been observed. AFLP markers were used to unravel hybridization and potential introgression events in this population. In total, 147 polymorphic loci remained after the process of peak selection. This dataset demonstrates hybrids originated from a cross between J. sabina and J. thurifera and suggests that back-cross at least to J. thurifera is possible. This study shed light on a rare case of hybridization of two Juniperus species with different ploidy levels in natural sympatric population.

The AFLP experiment was conducted according to Vos et al. (1995) with slight modifications. Around 200 ng of genomic DNA of each sampled individual was used for digestion and ligation. The PCR selective amplifications were elaborated using two different primer combinations given in the table 6 of the related manuscript. Each forward primer used in the selective amplification was labelled with 6-FAM modification at the 5’ extremity.

To test fragment detection repeatability, eight negative controls were included in the PCR and genotyping steps. In addition, 7 individuals randomly chosen (2 from J. thurifera and 5 from J. sabina) were repeated three times: two repetitions were done at the stage of extraction and kept all over the process, and the third one was included after the digestion-ligation step at PCR and genotyping steps.

AFLP migration profiles were analyzed using GeneMapper v.5 (Thermo Fisher Scientific) using the GS500 (-250) LIZ control to scale for fragment size.

Markers selection was conducted as following:

1. Peak selection and phenotype assignment was done using the default options in GeneMapper v.5 for AFLP data.
2. Peaks having a size between 60 bp and 500 bp were exclusively kept for the analysis.
3. Manual examination of peaks for each individual was done and individuals that displayed bad profiles (weak signals or strong background noise) were discarded from the analysis.
4. All peaks detected in negative controls were removed.
5. Bonin Error rate (Bonin et al. 2004) using repetitions of control samples were estimated and all peaks that had an Error > 0% were rejected.
6. Locus displaying alternative phenotypes (presence or absence of the detected peak in only one individual among the whole sample (singletons)) were also discarded.