Pedigrees have a long history in classical genetics, agronomics, evolutionary ecology, and ex situ captive breeding. Use of molecular techniques has expanded the variety of species for which pedigrees can be constructed. However, molecular pedigrees almost exclusively consider microsatellite loci, despite advances in high-throughput sequencing allowing development of genomic marker sets in nearly any organism. Here we generate a novel set of genomic SNPs derived from ddRAD sequencing in two populations of Weddell seals (Leptonychotes weddellii) and describe the diversity and differentiation between them. We then compare and contrast parentage assignment rates and accuracy in one population that has been the subject of long-term monitoring. Specifically, we consider pedigrees constructed using two sets of markers (microsatellites and SNPs), two pedigree construction software (CERVUS than Sequoia), as well as varying the groupings of candidate parents (either all individuals simultaneously, only individuals born before a focal year, or only individuals known to have survived to a focal year). ddRAD sequencing returned between 1568 and 3240 loci depending on whether both populations were considered simultaneously or individually. Parentage assignment rates were always higher using CERVUS than Sequoia, with the latter at times either not assigning parentage or creating “inferred parents”. In all cases, “polarizing” the datasets (e.g., including year of birth) significantly improved assignments. This represents one of the first direct comparisons of pedigree construction using different markers in the same set of individuals, and the SNPs described here will be a resource for continued pedigree construction, and future research in Weddell seals.

谱系在经典遗传学、农学、进化生态学和非原生境圈养育种方面有着悠久的历史。分子技术的使用扩大了可构建谱系的物种种类。然而，尽管高通量测序的进步允许在几乎任何生物体中开发基因组标记集，但分子谱系几乎只考虑微卫星位点。在这里，我们从两个威德尔海豹种群（Leptonychotes weddellii) 并描述它们之间的多样性和差异性。然后，我们在一个长期监测的对象中比较和对比亲子分配率和准确性。具体来说，我们考虑使用两组标记（微卫星和 SNP）、两个谱系构建软件（CERVUS 而不是 Sequoia）构建的谱系，以及改变候选父母的分组（所有个体同时，仅在焦点年份之前出生的个体，或仅已知存活至焦点年份的个体）。ddRAD 测序返回 1568 到 3240 个基因座，具体取决于是同时考虑还是单独考虑两个群体。使用 CERVUS 的亲子分配率总是高于红杉，后者有时不分配亲子关系或创建“推断的父母”。在所有情况下，“极化”数据集（例如，包括出生年份）显着改善了分配。这是在同一组个体中使用不同标记对谱系构建的首次直接比较之一，此处描述的 SNP 将成为继续谱系构建和威德尔海豹未来研究的资源。