Genotyping
The CHIIMP v 0.3.1 (Barbian et al., 2018) pipeline was used on reads filtered with CutAdapt to generate genotypes across all replicates using identical parameters. Each genotype was called if there were minimally 5 reads (counts.min =5), alleles were identified only if the number of recovered reads constituted minimally 5% of the filtered reads (fraction.min = 0.05), and all loci were given a length buffer of 20 bp. The output from CHIIMP not only identified the most likely allele calls, but also provided data on whether or not the sample had PCR associated stutter, PCR artifacts removed and more than two prominent sequences. Data for the individually library prepared samples, the pooled data and a bioinformatically pooled dataset, where the 2-3 replicates of each individually prepared reads were concatenated together, were put through quality filtering in order to evaluate the possibility of low coverage of the library preparations leading to erroneous genotypes. All of the recovered genotypes were manually compared in order to evaluate any patterns across the tested samples.