Department of Ecology and Evolutionary Biology, University of Michigan, 1109 Geddes Avenue, Ann Arbor, MI 48109-1079, USA.
Department of Ecology and Evolutionary Biology, University of Michigan, 1109 Geddes Avenue, Ann Arbor, MI 48109-1079, USA.
Trends Ecol Evol. 2018 Jun;33(6):390-398. doi: 10.1016/j.tree.2018.03.010. Epub 2018 Apr 21.
The development of process-based probabilistic models for historical biogeography has transformed the field by grounding it in modern statistical hypothesis testing. However, most of these models abstract away biological differences, reducing species to interchangeable lineages. We present here the case for reintegration of biology into probabilistic historical biogeographical models, allowing a broader range of questions about biogeographical processes beyond ancestral range estimation or simple correlation between a trait and a distribution pattern, as well as allowing us to assess how inferences about ancestral ranges themselves might be impacted by differential biological traits. We show how new approaches to inference might cope with the computational challenges resulting from the increased complexity of these trait-based historical biogeographical models.
基于过程的概率模型在历史生物地理学领域的发展，通过将其建立在现代统计假设检验的基础上，彻底改变了这一领域。然而，这些模型大多忽略了生物学差异，将物种简化为可互换的谱系。在这里，我们提出将生物学重新纳入概率历史生物地理学模型的理由，以便更广泛地探讨生物地理过程的问题，而不仅仅是祖先范围的估计或特征与分布模式之间的简单相关性，同时也使我们能够评估关于祖先范围本身的推断可能如何受到不同生物特征的影响。我们展示了新的推断方法如何应对这些基于特征的历史生物地理学模型复杂性增加所带来的计算挑战。
