Pablo de Olavide University, Environmental Science Faculty, Carretera, Utrera km-1, 41013-Sevilla, Spain.
Am J Bot. 2010 Sep;97(9):1555-73. doi: 10.3732/ajb.1000115. Epub 2010 Aug 26.
The use of continuous morphological characters in taxonomy is traditionally contingent on the existence of discrete diagnostic characters. When plant species are the result of recent divergence and gene flow and/or hybridization occur, the use of continuous morphological characters may help in species identification and delimitation. Between nine and 15 species have been recognized in the last treatments of Hedera. The recent divergence of the species and the involvement of allopolyploidization as the main force in this process may have greatly impeded the establishment of clear limits and contributed to multiple taxonomic proposals. •
A multivariate statistical decision-making procedure was applied to 56 quantitative morphological characters and 602 specimens to identify and delimit Hedera species under the general lineage concept. Species' exclusive genetic ancestry was evaluated with the genealogical sorting index from the Bayesian inference trees of 30 Hedera ITS sequences. •
The decision-making procedure allowed recognizing 12 species and two groups (stellate and scale-like trichome groups) in Hedera and provided statistical support for making decisions about long-standing taxonomic controversies. Common ancestry was detected for the populations of three species even in the absence of the species monophyly. •
Quantitative variation supports discrete variation and provides statistical support for the taxa recognized in some recent proposals of Hedera. The need of explicit analysis of quantitative data are claimed to reduce taxonomic subjectivity and ease decision-making when qualitative data fail.
传统上，在分类学中使用连续形态特征取决于离散诊断特征的存在。当植物物种是最近分化的结果，并且发生基因流和/或杂交时，使用连续形态特征可能有助于物种鉴定和划分。在过去对常春藤属的处理中，已经识别出了 9 到 15 个物种。物种的最近分化以及多倍体化作为这一过程的主要力量，可能极大地阻碍了明确界限的建立，并促成了多种分类建议。
应用多元统计决策程序对 56 个定量形态特征和 602 个标本进行分析，根据广义谱系概念识别和划定常春藤属的物种。通过对 30 个常春藤 ITS 序列的贝叶斯推断树中的系统发生分类指数，评估了物种的独特遗传渊源。
决策程序允许识别出 12 个物种和两个群（星状和鳞片状毛状体群），并为解决长期存在的分类争议提供了统计支持。即使在没有物种单系性的情况下，也检测到了三个物种的种群的共同祖先。
定量变异支持离散变异，并为一些最近提出的常春藤属分类群提供了统计支持。需要对定量数据进行明确分析，以减少分类学的主观性，并在定性数据失败时简化决策。
