Comparative Genomic Analyses and a Novel Linkage Map for Cisco (Coregonus artedi) Provide Insights into Chromosomal Evolution and Rediploidization Across Salmonids

Abstract

<jats:p>Whole-genome duplication (WGD) is hypothesized to be an important evolutionary mechanism that can facilitate adaptation and speciation. Genomes that exist in states of both diploidy and residual tetraploidy are of particular interest, as mechanisms that maintain the ploidy mosaic after WGD may provide important insights into evolutionary processes. The Salmonidae family exhibits residual tetraploidy, and this, combined with the evolutionary diversity formed after an ancestral autotetraploidization event, makes this group a useful study system. In this study, we generate a novel linkage map for cisco (<jats:italic>Coregonus artedi</jats:italic>), an economically and culturally important fish in North America and a member of the subfamily Coregoninae, which previously lacked a high-density haploid linkage map. We also conduct comparative genomic analyses to refine our understanding of chromosomal fusion/fission history across salmonids. To facilitate this comparative approach, we use the naming strategy of protokaryotype identifiers (PKs) to associate duplicated chromosomes to their putative ancestral state. The female linkage map for cisco contains 20,292 loci, 3,225 of which are likely within residually tetraploid regions. Comparative genomic analyses revealed that patterns of residual tetrasomy are generally conserved across species, although interspecific variation persists. To determine the broad-scale retention of residual tetrasomy across the salmonids, we analyze sequence similarity of currently available genomes and find evidence of residual tetrasomy in seven of the eight chromosomes that have been previously hypothesized to show this pattern. This interspecific variation in extent of rediploidization may have important implications for understanding salmonid evolutionary histories and informing future conservation efforts.</jats:p>

Genotyping‐by‐sequencing illuminates high levels of divergence among sympatric forms of coregonines in the Laurentian Great Lakes

“…Filtered reads for each individual were aligned to create matching stacks with ustacks following guidelines suggested from empirical testing to avoid under‐ or over‐merging loci from RAD datasets (Paris, Stevens, Catchen, & Johnston, ; parameter flags: ‐‐disable‐gapped ‐m 3 ‐M 5 ‐H ‐‐max_locus_stacks 4 ‐‐model_type bounded ‐‐bound_high 0.05). A catalog of consensus loci built from 51 artedi sampled in Lake Huron used for the development of a cisco linkage map (Blumstein et al, ) was appended in cstacks with an additional 75 individuals in the C. artedi species complex from across the Great Lakes, including 19 fish used in the current study: five hoyi and four kiyi from the Apostle Islands and five each of nigripinnis ‐like and zenithicus from various locations across Lake Superior (parameter flags: ‐n 3 ‐p 6 –disable_gapped). Locus stacks for each individual were matched to the catalog using sstacks (parameter flag: ‐‐disable_gapped), data were oriented by locus in tsv2bam , and reads were aligned to loci and SNPs were called with gstacks .…”

“…SNPs genotyped in greater than 30% of individuals (parameter flag: ‐r 0.3) were exported with the subprogram populations in genepop and variant call format (vcf) files. We isolated SNP calling to single‐end reads to ensure equal coverage across sequences, but paired‐end assemblies for each locus are available from Blumstein et al ().…”

“…We used a p‐crit of 0.05 for the RAD dataset (Waples, Larson, & Waples, ) and 0.02 for the microsatellite dataset (Waples & Do, ). For the RAD dataset, only comparisons between sampled loci that were found on different linkage groups (LGs) of the cisco linkage map (Blumstein et al, ) were included to correct for physical linkage (Waples et al, ). N e calculations using the linkage disequilibrium method can be biased slightly downward when individuals from multiple cohorts are included in the sample due to a slight Wahlund effect (7% downward bias on average; Waples, Antao, & Luikart, ).…”

“…We examined genetic differentiation across the genome by pairing our data with the artedi linkage map constructed by Blumstein et al (). Catalog IDs were identical between the current study and linkage map; therefore, no alignment step was needed to compare loci.…”

“…The main goal of our study was to employ genomic methods to improve our understanding of genetic variation among these forms. Specifically, we (1) examined genetic differentiation and diversity among putative forms of cisco, (2) compared the performance of SNPs and microsatellites in this system, and (3) leveraged a newly built cisco linkage map (Blumstein et al, ) to investigate adaptive divergence among forms. In order to remove the potentially confounding factor of distinguishing spatial genetic structure from form‐based genetic structure, we focused on a single region in Lake Superior where multiple forms of cisco are found in sympatry, the Apostle Islands.…”