Because the true phylogeny for most organisms can never be known with certainty to us, congruence between/among inferred phylogenies from two or more independent data sets provides strong evidence for the reliability of phylogenetic trees. Therefore, there has been increasing numbers of data sets obtained and analysed from the same set of taxa for phylogeny reconstruction in recent years. However, it is also possible that different data sets may imply different stories rather than the same evolutionary history. In this case, there has been growing concern over how to integrate different data sets in phylogenetic analysis. Three general approaches have been advocated for obtaining phylogenies based on two or more data sets. First, the total evidence or character congruence approach involves pooling all the available data into a single data matrix and finding a globally parsimonious solution. Second, the consensus or taxonomic congruence approach involves estimating the independent phylogeny from each data set and constructing a consensus tree based on several different consensus tree concepts. Third, the tree recoiling approach involves recoding the tree topology into binary characters and analysing the recoded data matrix. In this paper, I first introduce both logical and methodological backgrounds of the three approaches. Then three approaches are applied previously published phylogenetic data on two plant (Krigia and Fuchsia) and two animal groups (Arthropoda and Amauris). In addition, I discuss several common issues which can be encountered during combined approaches such as character swampping, complementary resolution, mosaic topology, pseudofossils, etc.
PDF Links
Full text via DOI
Download Citation 
