Data from: Bayesian inference of a historical bottleneck in a heavily exploited marine mammal

Emerging Bayesian analytical approaches offer increasingly sophisticated means of reconstructing historical population dynamics from genetic data, but have been little applied to scenarios involving demographic bottlenecks. Consequently, we analysed a large mitochondrial and microsatellite dataset from the Antarctic fur seal Arctocephalus gazella, a species subjected to one of the most extreme examples of uncontrolled exploitation in history when it was reduced to the brink of extinction by the sealing industry during the late eighteenth and nineteenth centuries. Classical bottleneck tests, which exploit the fact that rare alleles are rapidly lost during demographic reduction, yielded ambiguous results. In contrast, a strong signal of recent demographic decline was detected using both Bayesian skyline plots and approximate Bayesian computing, the latter also allowing derivation of posterior parameter estimates that were remarkably consistent with historical observations. This was achieved using only contemporary samples, further emphasizing the potential of Bayesian approaches to address important outstanding problems in conservation and evolutionary biology.新兴贝叶斯分析方法正为从遗传数据中重建历史种群动态提供日益精密的技术手段，但此类方法在涉及种群瓶颈的研究场景中应用甚少。为此，我们针对南极海狗（Arctocephalus gazella）的大型线粒体与微卫星数据集开展了分析——该物种曾在18至19世纪末的海豹捕猎产业影响下遭到无节制开发，种群一度濒临灭绝，属于史上最极端的此类开发案例之一。经典瓶颈测试借助“稀有等位基因在种群缩减过程中会快速丢失”这一特性开展，却得到了模棱两可的结果。与之形成鲜明对比的是，通过贝叶斯天际线图（Bayesian skyline plots）与近似贝叶斯计算（approximate Bayesian computing）两种方法，均检测到了近期种群衰退的强烈信号；其中近似贝叶斯计算还推导得出了与历史观测记录惊人一致的后验参数估计值。本研究仅使用当代样本便达成了上述分析目标，进一步凸显了贝叶斯方法在解决保护生物学与进化生物学领域重要悬而未决问题方面的潜力。