Sex Chromosome Identification
Although karyotype analysis has been conducted in several Nebriaspecies (Serrano & Galián 1998; Serrano & Yadav 1984), the sex chromosome system has not been identified. However, either XY and XO systems are present in other carabid beetles (Serrano & Yadav 1984), so we employed two computational approaches to identify the X chromosome (Palmer et al. 2019). First, because female individuals are expected to have two copies of the X chromosome and male individuals only have one copy, we examined sex-specific coverage to distinguish X chromosome scaffolds. We used a published dataset (Weng et al.2020) of genotyping by sequencing (GBS) reads from 59 N. riversiindividuals (BioProject PRJNA645878, specimen data listed inTable S1 ), comprising 34 females and 25 males from three sites, to calculate the sequencing depth across the genome. Single-end GBS reads were mapped to the genome using Bwa-Mem v0.7.17.4 to generate the mapped bam files (Li 2013) and Samtools v1.10.03 was used to calculate the mean depth of coverage for each scaffold (Liet al. 2009). Welch’s t-tests were conducted for each contig to test whether mean coverage of male individuals was half that of female individuals. All contigs with p-value < 0.05, and/or showing a ratio of male/female coverage between 0.45 and 0.55, were considered X chromosome candidates. Second, we further processed the GBS data to identify single nucleotide polymorphisms (SNPs) among the candidate X chromosome scaffolds. We added the read group to the bam files and sorted them using Picard v2.23.3 (Picard 2018), then called SNPs using HaplotypeCaller in GATK v.4.1.8.1 (McKenna et al. 2010; Poplin et al. 2017). We examined whether scaffolds contained SNPs that were heterozygous in males, as male individuals are expected to carry a single copy of X chromosome and thus should show single allele in those loci under an ideal situation without sequencing errors. In order to keep the high quality SNPs, we first filtered by read depth (>10x) and missing data, requiring genotype calls in at least half of female or male samples. The SNPs that passed the filter were used to calculate the mean read depths for males and females, and the scaffolds carrying SNPs with mean read depths ratio of male/female between 0.4 and 0.6 were marked as candidates. We repeated this for each population from the three sample sites (Conness Lake, Donohue Pass, and Lyell Peak; for more information about collecting information see: Weng et al. 2020) and all samples together, and the contigs that identified in all four data sets were considered X chromosome candidates. Finally, all X chromosome candidates were screened for the presence of heterozygote SNPs in males, and only those scaffolds that had a heterozygote rate less than the standard sequencing error rate of 0.01 were retained as X chromosome candidates.