Paraleucophenga undetermined

Zhao, Feng, Gao, Jian-Jun & Chen, Hong-Wei, 2009, Taxonomy and molecular phylogeny of the Asian Paraleucophenga Hendel (Diptera, Drosophilidae), Zoological Journal of the Linnean Society 155 (3), pp. 615-629 : 618

DIAGNOSING PARALEUCOPHENGA SPECIES USING

DNA SEQUENCES

All the ND2 sequences, including those of the outgroup taxa, were aligned using the ClustalW ( Thompson, Higgins & Gibson, 1994) method, and were then modified manually. The resulting alignment was used for diagnostic character (nucleotide position) selection, the DNA substitution model test, as well for phylogenetic reconstruction. Groups of nucleotide positions were selected as characters to diagnose different taxa (pairs of species or species clusters) in the genus Paraleucophenga . For this, less degenerated positions showing amino acid replacement between (but not within) the two taxa compared were preferentially considered. A ‘molecular’ key using the selected positions as diagnostic characters was elaborated to distinguish between the Paraleucophenga species for which ND2 sequences were available.

PHYLOGENETIC ANALYSIS

A DNA substitution model was selected using Modeltest 3.7 ( Posada & Crandall, 1998), and phylogenetic analyses were performed with the maximum parsimony (MP) and maximum likelihood (ML) methods using PAUP*4.0b10 ( Swofford, 2003), and with the Bayesian inference (BI) method using MrBayes 3.1 ( Ronquist & Huelsenbeck, 2003). Sites with gaps were treated as missing data in the MP analysis, but were completely deleted in the ML and BI analyses. Calculation of basic statistical quantities of DNA sequences were implemented in MEGA 3.1 ( Kumar, Tamura & Nei, 2004). The heuristic search of MP trees was performed with initial trees obtained by 100 replicates of random addition, and branch swapping using the tree bisection and reconnection (TBR) algorithm. The K81uf + G (two transversion parameters, model 1, unequal frequencies, and a gamma-distributed substitution rate across sites) model was selected for the ML analysis, with parameters assigned as follows: base frequencies of A, C, G, and T are 0.3695, 0.0974, 0.0584, and 0.4746, respectively; substitution rates for the classes A–C, A–G, A–T, C–G, C–T, and G–T were 1.0000, 12.7822, 1.9286, 1.9286, 12.7822, and 1.0000, respectively; the proportion of invariable sites = 0; gamma distribution shape parameter = 0.2820. For the Bayesian analysis with MrBayes 3.1 ( Ronquist & Huelsenbeck, 2003), a site-specific rate model was used, allowing the first codon position to have six substitution types, whereas the second and third codon positions were only allowed to have two substitution types. For all of the codon positions, a gamma-distributed rate was assumed. For the MP and ML analyses, the bootstrap percentages (BPs; 1000 replicates) were calculated to evaluate the node confidences.