Population genomics of the killer whale; SNP discovery towards the assessment of population structure at functional genes

自从 20 世纪 70 年代加拿大的一个研究项目首次展示了如何使用照片来追踪区域性虎鲸种群，以识别和跟踪个体鲸鱼的关联和活动以来，越来越明显的是，该物种为增进我们对关键进化和生态过程的理解提供了巨大的潜力。尽管社会群体在一个季节内移动数千公里，具有长距离分散的巨大潜力，但已确定的关联个体种群（分为20-50个个体的社会单位）在小地理范围内的中性标记上存在遗传差异。明显的机制是依赖当地或季节性栖息地资源，特别是猎物资源。通过这种方式，可以根据溯河产卵鲑鱼猎物的迁徙方向（向河口以北或以南迁徙）来区分食鱼种群，并且即使在同源的情况下，也可以将食鱼种群与以海洋哺乳动物为食的种群隔离开来。存在与地理距离相关的分化模式（正如许多物种所见），但仅限于生态型内。生态型之间的分化基于生物因素（特别是与觅食和社会结构相关的行为）。尽管区域种群的有效规模可能很小，但有时存在可能与资源开发相关的物理差异（如在各种其他翠雀物种中所见，特别是与口腔形态相关），这表明当地的适应。在这项研究中，我们将识别明显处于选择之下的基因，并尽可能将它们与生态特征联系起来，特别是不同生态型种群生活史的已知方面。一个关键的交付成果将是一套广泛的 SNP 标记，其中一些被确定为可能正在选择中，将应用于广泛的人群筛查计划的后续研究。作为 SNP 发现过程的一部分，本研究将基于这些 SNP 标记进行初步评估。这是通过所选择的方法实现的。第一步将对一个样本的全基因组进行测序，达到足以用作参考支架的质量水平（通过商业提供商 Eurofins MWG/Operon 完成）。然后，这将用于促进 RAD-Tag 测序工作和相关的生物信息学（将在 CGR Liverpool 合作伙伴 Neil Hall 的监督下进行）。然后，达勒姆的 PI 将与 Hall 教授合作并在 RA 的协助下对 RAD 标签数据进行分析。 RAD 标签序列将允许识别 SNP 标记，参考基因组将允许识别它们在虎鲸基因组中的位置和连锁。对来自 5 个种群的最初 150 头鲸鱼的分析将为选择信号和分化模式的高分辨率评估提供初步数据。将应用贝叶斯和多元回归方法来解释环境变量背景下的遗传变异，并评估哪些变量解释了遗传方差的最大比例。总而言之，这些分析将增进我们对物种内部和物种之间产生生物多样性的进化过程的理解，促进更有效的生物多样性保护计划，并为进一步推进这些目标的未来研究提供原材料。

Since a Canadian research project in the 1970s first showed how the regional killer whale population could be tracked over time using photographs to identify and follow the associations and movement of individual whales, it has become increasingly apparent that this species offers great potential for advancing our understanding of key evolutionary and ecological processes. Despite their tremendous potential for long-distance dispersion, with social groups travelling 1000s of km in a season, identified populations of associating individuals (divided into social units of 20-50 individuals) are genetically differentiated at neutral markers over small geographic scales. The apparent mechanism is dependence on local or seasonal habitat resources, and in particular, prey resources. In this way picivorous populations can be separated by the direction of migration in their anadromous salmonid prey (migrating north or south of an estuary), and picivorous populations can be isolated from marine-mammal-eating populations even in sympatry. There is a pattern of differentiation associated with geographic distance (as seen for many species), but only within an ecotype. Differentiation among ecotypes is based on biological factors (especially behaviour associated with foraging and social structure). Although the effective size of regional populations is likely small, there are sometimes physical differences that may be associated with resource exploitation (as seen in various other delphinid species, especially associated with buccal morphology), suggesting local adaptation. In this study we will identify genes that are apparently under selection, and associate them where possible with ecological characteristics, especially known aspect of the life history of the different ecotype populations. A key deliverable will be an extensive set of SNP markers, with some identified as likely under selection, to be applied in a follow-on study to an extensive program of population screening. A preliminary assessment based on these SNP markers will be undertaken in the current study as part of the SNP discovery process. This is made possible by the chosen methodology. The initial step will be the sequencing of the full genome for one sample to a quality level sufficient to use as a reference scaffold (to be done through the commercial provider, Eurofins MWG/ Operon). This will then be used to facilitate the RAD-Tag sequencing work and the associated bio-informatics (to be undertaken under the supervision of partner Neil Hall at CGR Liverpool). The RAD-Tag data will then be analysed by the PI in Durham in collaboration with Prof. Hall and with the assistance of the RA. The RAD-Tag sequences will permit the identification of SNP markers, and the reference genome would permit identification of their location and linkage within the killer whale genome. Analysis of an initial 150 whales from 5 populations will provide the preliminary data on signals for selection and a high-resolution assessment of patterns of differentiation. Bayesian and multiple regression methods will be applied to allow interpretation of genetic variation in the context of environmental variables, and to assess which variables explain the largest proportion of the genetic variance. Taken together these analyses will advance our understanding of the evolutionary processes that generate biodiversity within and among species, facilitate more effective programs of biodiversity conservation, and provide the raw material for future studies that will advance these objectives further.