期刊
MOLECULAR PHYLOGENETICS AND EVOLUTION
卷 66, 期 3, 页码 603-616出版社
ACADEMIC PRESS INC ELSEVIER SCIENCE
DOI: 10.1016/j.ympev.2012.09.024
关键词
Cyprinidae; Nuclear genes; Phylogeny; Mixed model Bayesian analyses
资金
- National Basic Research Program of China (973 Program) [2010CB126302]
- National Natural Science Foundation of China [31090254, U1036603]
- Chinese Academy of Sciences [KSCX2-EW-Q-12]
- USA NSF EF [0431326, DEB-1021840, DBI-0956370]
Despite many efforts to resolve evolutionary relationships among major clades of Cyprinidae, some nodes have been especially problematic and remain unresolved. In this study, we employ four nuclear gene fragments (3.3 kb) to infer interrelationships of the Cyprinidae. A reconstruction of the phylogenetic relationships within the family using maximum parsimony, maximum likelihood, and Bayesian analyses is presented. Among the taxa within the monophyletic Cyprinidae, Rasborinae is the basal-most lineage; Cyprinine is sister to Leuciscine. The monophyly for the subfamilies Gobioninae, Leuciscinae and Acheilognathinae were resolved with high nodal support. Although our results do not completely resolve relationships within Cyprinidae, this study presents novel and significant findings having major implications for a highly diverse and enigmatic clade of East-Asian cyprinids. Within this monophyletic group five closely-related subgroups are identified. Tinca tinca, one of the most phylogenetically enigmatic genera in the family, is strongly supported as having evolutionary affinities with this East-Asian clade; an established yet remarkable association because of the natural variation in phenotypes and generalized ecological niches occupied by these taxa. Our results clearly argue that the choice of partitioning strategies has significant impacts on the phylogenetic reconstructions, especially when multiple genes are being considered. The most highly partitioned model (partitioned by codon positions within genes) extracts the strongest phylogenetic signals and performs better than any other partitioning schemes supported by the strongest 2 Delta ln Bayes factor. Future studies should include higher levels of taxon sampling and partitioned, model-based analyses. (C) 2012 Elsevier Inc. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据