TABLE 2.

Filtration steps of ddRADseq and lcWGS datasets for sample and single nucleotide polymorphisms (SNPs). Each row is a filtration step targeting either samples or SNPs for selection or removal, and provides retention counts combining a given step and filtering steps in preceding rows. For the lcWGS dataset, prior to pruning for linked sites, mean SNP depth was 4.45×, while site missingness was 2.6% (Figure S12).

Filtration steps	Target	ddRADseq			lcWGS
		SNPs	Loci	N	SNPs	N
Initial samples	—	—	—	755	—	351
Reads mapping ≥ 96%	Sample	—	—	746	71,851,401	344
Mean depth coverage ≥ 5× (post gstacks)	Sample	—	2,138,443	680	—	—
Located outside repetitive elements	SNP	—	—	—	41,966,793	344
Biallelic	SNP	—	—	—	41,232,252	344
Located further than 5 bp of an indel	SNP	—	—	—	40,867,199	344
Not an indel (SNPs only)	SNP	—	—	—	40,312,568	344
Minor Allele Frequency (MAF) > 1% or SNP detected in > 25% samples	SNP	136,771	88,544	680	—	—
Read depth > 15× and < 29×	SNP	106,757	69,936	680	—	—
SNP < 10% missing data and sample < 30% (ddRADseq) or < 10% (lcWGS) missing loci	SNP, sample	92,115	62,195	678	39,291,750	341
Mean read depth > 5×	Sample	92,115	62,195	677	—	—
MAF > 5%	SNP	—	—	—	3,000,827	341
Not linked to sex a	SNP	—	—	—	2,946,667	341
Observed heterozygosity < 60%	SNP	91,967	62,069	677	2,922,691	341
Sequencing plates effect	SNP	90,117	61,159	677	—	—
Sex‐linked and located within repetitive elements	SNP	88,433	60,102	677	—	—
Relatedness (Φ < 0.25)	Sample	88,433	60,102	638	2,922,691	341
One SNP per locus	SNP	60,102	60,102	638	—	—
MAF > 5% and < 5% missing data		26,019	26,019	638	—	—
Unlinked loci (r 2 < 0.25), 50 kbp sliding window	SNP	—	—	—	845,731	340 b
Final dataset complete		26,019	26,019	638	845,731	340 b
Final dataset without outliers		24,709	24,709	638

^a The Y chromosome was not compiled into the original VCF file and sites on the X chromosome were removed.

^b One individual was removed from the final dataset due to conflicting genetic signal in replicate samples.