Computational protocol: Conserved community structure and simultaneous divergence events in the fig wasps associated with Ficus benjamina in Australia and China

Similar protocols

Protocol publication

[…] New sequences were checked using the chromatogram visualisation software Finch TV Version 1.4.0 and were edited and aligned using BioEdit []. Final adjustments were made by eye. Bayesian methods were employed to construct phylogenies for each marker using MrBayes []. The best model of nucleotide substitution for each gene was determined using likelihood ratio tests with MrModeltest in PAUP* []. For both markers the general time reversible (GTR + I + R) model of evolution was chosen.Species richness was calculated using the barcoding criteria of monophyletic lineages of individuals (here for both markers) with at least one other conspecific within a commonly used cut-off threshold of 3% K2P (Kimura-2-parameter measure of pairwise genetic distance []) pairwise difference in the COI marker [], or the 95% confidence interval of all intraspecific distances []. COI K2P distances were calculated [in Mesquite; ] for congeneric sister clades of wasps from different continents. We then further tested these species hypotheses on COI sequences using the jMOTU software package [] that is designed to cluster sequences according to a range of cut-off values—an inflection point in jMOTU outputs purportedly shows the barcoding gap. [...] Split dates between congeneric species were estimated from COI data using Bayesian dating analyses in BEAST v.1.7.1 []. As mitochondrial DNA (mtDNA) evolves quickly and relationships among deeper nodes lose signal (e.g. within a multi-family phylogeny), we ran individual analyses on the following genera: Eupristina, Walkerella, Philotrypesis, Sycobia and Sycophila, as well as on the Sycoryctini tribe (Sycoscapter and Sycorycteridea). As no fossil calibration points are available, a molecular clock, estimated for Asian fig wasps (1.9% pairwise divergence per million years; [, ]), was imposed. Bayes factors (BF) were calculated using Raftery’s [] logarithmic scale (2loge(BF)) to detect the most suitable models of substitution, molecular clock and speciation (where values of 2–5 on this scale are considered positive evidence, > 5 is strong evidence, and > 10 is very strong evidence). Two independent runs of 120 million iterations were run for each species separately. Genealogies and model parameters were sampled every 6000 iterations and 95% highest posterior densities (HPD) were employed as credibility intervals. The first 2000 trees of each run were discarded as the “burn-in” period. […]

Pipeline specifications