Marker Evaluation
Across all sequenced samples, individuals had on-target reads, those reads corresponding to amplified target fragments, ranging from 1 to 1.71 million, with a mean of 221k reads, corresponding to OT read percentages (OT reads/total reads) of mean 55% (~0 to 86%). The allele ratio plots revealed that most loci conformed to expectation of tetrasomic segregation (Number of loci with scores of 1, 2, 3, and 4: 97, 133, 56, and 39, respectively). Moreover, score 4 loci, which often did not exhibit five clear genotype categories, did not clearly conform to expectations of octosomy either (Supplemental Figure 2). Rather, there was a moderate and significant correlation (p<0.001) between these allele ratio scores and both percent Mendelian incompatibilities in parent-offspring comparisons (R2 = 0.59) and ranked likelihood of octosomic inheritance (R2 = 0.36)(Figure 1). This suggests that most or all of these SNP loci exhibit tetrasomic inheritance, though some may be affected by allelic bias or co-amplification of non-target loci, the latter of which may reflect homeologs in some cases. In addition, even the score 4 loci had a mean percent Mendelian incompatibility of 0.08, suggesting that despite some noise or skew, most of these loci still exhibit relatively reliable genetic signal.
While white sturgeon are ancestral octoploids (Drauch Schreier et al., 2011), it remains unclear what the contemporary meiotic segregation patterns are for this and other sturgeon. Indeed, even studies using similar data sources, such as microsatellite genotype and inheritance patterns, have come to conflicting conclusions (Drauch Schreier et al., 2011; Ludwig, Belfiore, Pitra, Svirsky, & Jenneckens, 2001). Perhaps not surprisingly, Du et al. (2020) found the sterlet genome exhibited both diploid and tetraploid characteristics, with entire segments of some chromosomes diploidized (homeologs lost or diverged), and others still syntenous across large stretches. Unfortunately, microsatellite markers may be equally subject to prolonged homeology, and are thus only helpful to resolve the question of segregation in so far as primers are exclusive to a single segregating locus rather than amplifying multiple separately segregating homeologs (Clark & Schreier, 2017). Moreover, this ambiguity as to how many homeologs are being amplified by any particular marker has complicated efforts to utilize genetic data from white sturgeon to aid conservation. Results from the present study suggest that a significant part of the white sturgeon genome is functionally tetraploid, although our bioinformatic process, which excluded variants with higher divergence and/or read depth, could have biased our survey to parts of the genome further along the process of tetraploidization (but see A. L. Van Eenennaam et al., 1998). Nonetheless, the observation that almost all of the molecular markers presented herein are clearly tetrasomic in 4N/8N individuals makes their application to conservation genetic analyses more straightforward, as demonstrated below.