Applying integrative multispecies coalescent models to tackle the long-standing problem of species delimitation in allopatrically distributed populations

基于多物种聚结模型（MSC）是解决物种划分这一关键问题的一种方法。但最近的研究表明，这些模型倾向于过度分割异域种群，因为它们受到种群结构而不是物种分化的单独影响。我们建议利用新热带鱼科的异域种群来解决物种划分问题，该科的分类学和系统学在过去几年中一直在变化，最近也被我们描述了几个新物种。这个科提供了一个有趣的系统来研究一般的物种划分问题，因为它包含了两个描述的物种，它们的种群发生在已经隔离了几百万年的流域中（即，可以忽略的基因流），而一些物种对在流域中是异栖的（即，潜在的基因流）。此外，这些鱼很容易在可控的条件下繁殖，允许进行特定的杂交来测试生殖隔离是否确实在单独的同源谱系之间进化，生殖隔离是根据生物物种概念完成物种形成的主要要求。我们的目标是整合全基因组重测序数据（脊椎动物基因组计划即将发布该家族的参考基因组）和表型数据（bbb70形态特征和几何形态计量学），应用结合这些数据集的新发展的基于聚结的综合分类方法，并使用基因组数据进行贝叶斯参数估计，并结合启发式谱系差异指数。然后，我们将通过将其与经验确定的物种对之间的生殖隔离程度进行比较，验证基于msc的物种划分的输出。最后，我们将划分我们的分子数据来研究基因座选择对基于msc的物种划分的影响。无尾鱼科是一个相对较小的鱼类科，只有大约20个分类群，它提供了一个可管理的、但在进化上非常有趣的测试案例，人们可以比较并详尽地测试这些新方法，以帮助识别、分类和描述新物种，这些新方法可以在以后扩大到其他物种丰富的谱系。

Multispecies coalescent-based models (MSC) were proposed as a solution for the crucial problem of species delimitation. But recent work suggested that these models tend to over-split allopatric populations as they are affected by population structure rather than species divergence alone. We propose to address the problem of species delimitation using allopatric populations of a Neotropical fish family, the Anablepidae, in which taxonomy and systematics have been in flux in the last years and several new species have been recently described, also by us. This family provides an interesting system to study general species delimitation problems as it contains both described species whose populations occur in drainages that have been isolated for several million years (i.e., negligible gene flow), whereas some pairs of species are parapatric within drainages (i.e., potential for gene flow). Additionally, these fish are easy to breed under controlled conditions, allowing to do specific crosses to test if reproductive isolation, the major requirement for the completion of speciation according to the Biological Species Concept, has indeed evolved between separate coalescent lineages. We aim to integrate whole-genome resequencing data (the Vertebrate Genome Project will soon release a reference genome for the family) and phenotypic data (> 70 morphological characters and geometric morphometrics) to apply newly developed coalescent-based integrative taxonomic approaches that combine these data sets, as well as use genomic data to conduct Bayesian parameter estimation combined with heuristic genealogical divergence index. Then, we will validate the output of the MSC-based species delimitations by contrasting it to the empirically determined degree of reproductive isolation among species pairs. Finally, we will partition our molecular data to study the impact of locus selection on MSC-based species delimitation. The familiy Anablepidae constitutes a relatively small family of fish with only about 20 taxa and provides a manageable, but evolutionary highly interesting test case in which one can compare, and exhaustively test, these new methods to aid in the identification, classification and description of new species that can be later scaled up to other more species-rich lineages.