Bioinformatic analysis
The data analysis was performed using the software package CHIIMP v.3.0.0 (Barbian et al. 2018). The raw data (FASTQ files) as well as all input files (config-file, sample-file, locus-attributes-file) are available in the online supplement resources. As our microsatellite panel included several di-repeat loci, which stutter more frequently than tetra-repeats, we increased the stutter count ratio to 0.70 (stutter.count.ratio_max: 0.7). We further implemented a broad range of possible allele lengths in the locus attributes by setting the length buffer to 100 bp. This ensured the inclusion of all tested species even if the allele sizes at a given locus varied between species according to the available reference genomes. The minimum number of reads per locus was set to 100 (counts.min: 100). All other parameters were set to default.
With the current version of CHIIMP, wobble positions in primer sequences cannot be accounted for. Hence, for loci with a wobble position in a primer sequence, alternative nucleotides of the wobble are erroneously recognized as different alleles. Moreover, the repeat motif needs to be specified in CHIIMP, but as repeat motifs can vary in the investigated species, correct (orthologous) reads remain unrecognized for some species if CHIIMP is fed with a wrong repeat motif. Due to these reasons, the output for all loci was checked manually and corrected if needed. Additionally, we screened the processed reads for the general level of amplification per locus and the occurrence of PCR artifacts (off target amplification, primer dimer, false primer pairings, etc.).