Species Delimitation in East African Cichlid Fish Adaptive Radiations

分类学是记录和理解生物多样性随时间变化的必要手段。为了更好地发现、命名和量化生物多样性，我们需要更好的模型和标准来有效地、可重复地将分类学与庞大的系统基因组数据集结合起来。直到最近，才有了基因组学和计算工具，使分类学家和系统学家能够为极其年轻的物种建立可靠的分类假设。对于更多的实验室来说，全基因组从头测序和重测序（WGS）在技术上和经济上都变得可行。这也适用于研究超多样化和通常极其年轻的东非慈鲷（基因组大小=< 1Gb）的适应性辐射。维多利亚湖的物种群特别年轻（可能在不到12,000年的时间内出现了500种），而马拉维湖的物种群（800-1000种，2-4 MYA）与坦噶尼喀湖更古老的谱系（250种，高达8-10MYA）相比，处于中等年龄。杂交、不完整的谱系分类和古代多态性的保留都是导致基因/等位基因树与物种树不一致的原因。这些现象被认为对年轻的物种来说尤其麻烦，但对坦噶尼喀湖的老物种来说问题不大。我们计划从三种辐射中系统地选择大约200个物种对的慈鲷基因组（每个物种25个个体，能够以5%的频率发现等位基因）进行重测序（参考基因组可用）。这些大量的序列数据也对如何阐明进化过程之间的关系以及遗传变异如何在谱系和物种之间分布提出了新的挑战。在基因组时代对新分类群的诊断将需要系统基因组学分析，以解释影响基因组结构、种间基因流动和物种形成的时间尺度之间的相互作用。利用整个基因组，我们计划对这些数据进行探索、划分和分析，以建立基于特定分子标记集的更客观和可重复的分类标准。WGS数据将被完整地分析，也将被分割成几个数据集，这些数据集将通过结构分析、主成分分析和贝叶斯统计方法来评估基因和物种树之间的一致性，目的是设计分类群诊断规则，当然，这些规则预计不仅适用于慈鲷。

Taxonomy is necessary for documenting and understanding changes in biodiversity through time. To better discover, name, and quantify organismal diversity, we need better models and criteria for how to effectively and reproducibly integrate taxonomy with huge phylogenomic data sets. Only very recently have genomic and computational tools became available that permit taxonomists and systematists to establish robust taxonomic hypotheses also for extremely young species. Whole-genome de novo sequencing and re-sequencing (WGS) is becoming technically and financially feasible for more laboratories. This applies also to research on hyper-diverse and often extremely young East African adaptive radiations of cichlid fishes (genome sizes =< 1Gb). The species flocks in Lake Victoria are particularly young (500 species arose in probably less than 12,000 years) whereas those of Lake Malawi (800-1000 species that are 2-4 MYA) are of intermediate age compared to the much older lineages of Lake Tanganyika (250 species that are up to 8-10MYA). Hybridization, incomplete lineage sorting, and the retention of ancient polymorphisms all contribute to gene/allele trees not agreeing with species trees. These phenomena are believed to be particularly onerous among the younger species, but less problematic for the older species of Lake Tanganyika. We plan to re-sequence (reference genomes are available) about 200 cichlid genomes of systematically chosen species pairs from each of the three radiations (25 individuals per species to be able to discover alleles at 5% frequency). These vast amounts of sequence data also cause novel challenges on how to elucidate the relationships between evolutionary processes and how genetic variation is distributed among lineages and species. The diagnoses of new taxa in the genomic age will demand phylogenomic analyses that account for the interplay between processes that affect genome architecture, interspecific gene flow, and the time-scales for speciation. Using entire genomes we plan to explore, partition, and analyze these data in an effort to develop more objective and reproducible taxonomic criteria based on particular molecular marker sets. The WGS data will be analyzed in full and also in several partitioned data sets whose congruence between gene and species trees will be evaluate through analyses including Structure, PCA, and Bayesian statistical methods, with the aim to devise rules for taxon diagnosis that, of course, are expected to apply not only for cichlids.