A Test of DNA Barcoding (a.k.a. DNA Taxonomy) and the Evolution of a Continental Fauna

这项研究首次尝试评估基于DNA的分类学方法，该方法结合了大规模测序、专家形态学-分类学知识和地理环境，其中包括几个谱系的大多数物种，而不像早期的研究只处理更遥远的物种。其结果将允许批判性地评估最近提出的批评要点。分类学被认为处于危机之中（Gaston & May, 1992; Godfray, 2002）。为了加快分类描述和知识传播的速度，提出了许多补救措施。Tautz等人（2003）提倡一种基于dna的方法，其中一个或几个基因区域作为物种普遍可获取的识别工具，Hebert等人（2003a, b）提出了一种单基因“条形码”方法，其中部分细胞色素c氧化酶I基因作为所有动物生命的识别系统。条形码是目前最广为宣传的方法。初步研究表明，DNA条形码表现良好（例如Hebert等人，2003b； Hebert等人，2004；Hogg和Hebert， 2004）。另一方面，发表了对这些方法的一般性批评，有时不幸出于政治动机（例如Lipscomb等人，2003;Will & Rubinoff, 2004），而不是提供严格的测试。我们的目标是对澳大利亚潜水甲虫动物群进行测序和形态学分类，包括几个特有的辐射，评估不同的聚类方法来描绘可以理想地代表物种的单倍型群。在这项研究结束时，我们期望有一个清晰的画面，什么序列数据可以和不能做，以帮助分类。

This study is the first attempt to evaluate DNA based approaches to taxonomy combining larger-scale sequencing, expert morphological-taxonomic knowledge and a geographical setting, which includes most species of several lineages ¿ unlike earlier studies which treated more distantly related species only. Its outcomes will allow to critically evaluate the points of criticism raised recently. Taxonomy was said to be in crisis (Gaston & May, 1992; Godfray, 2002). Many remedies have been proposed to accelerate the rate of taxonomic description and dissemination of knowledge. Tautz et al. (2003) advocated a DNA-based approach where one or several gene regions serve as universally accessible identification tool for species, Hebert et al. (2003a, b) proposed a single-gene ¿barcode¿ approach where parts of the cytochrome c oxidase I gene serve as an identification system for all animal life. Barcoding currently is the most widely advertised approach. Preliminary studies suggested that DNA barcoding performs well (e.g. Hebert et al., 2003b; Hebert et al., 2004; Hogg & Hebert, 2004). On the other hand, general criticism of the methods was published, sometimes unfortunately politically motivated (e.g. Lipscomb et al., 2003; Will & Rubinoff, 2004) rather than providing a rigorous test. We aim to sequence and morphologically classify the Australian diving beetle fauna comprising several endemic radiations, evaluating different clustering methods to delineate groups of haplotypes which could ideally represent species. At the end of this study we expect a clear picture of what sequence data can and can not do to aid taxonomy.