Computational protocol: Oceanic dispersal barriers, adaptation and larval retention: an interdisciplinary assessment of potential factors maintaining a phylogeographic break between sister lineages of an African prawn

Similar protocols

Protocol publication

[…] Tissue samples from 35 mudprawns each were collected in four Wild Coast estuaries by either breaking off a specimen's right first pereipod (large specimens) or preserving complete individuals (small specimens) in preservation medium containing 70% ethanol and 30% TE buffer. DNA extraction, amplification of a portion of the COI gene and sequencing were performed as described previously using crustacean forward primer 5'-TCA ACA AAT CAY AAA GAY ATT GG-3' and decapod reverse primer 5'-AAT TAA AAT RTA WAC TTC TGG-3' []. The sequences generated in this study were submitted to GenBank (accession numbers FJ416903–FJ417042). The program ARLEQUIN 3.1 [] was used to calculate each population's haplotype and nucleotide diversity based on pairwise differences, and an exact test of sample differentiation based on haplotype frequencies [] was performed for pairs of populations. To visualise which lineages were represented in each of the estuaries sampled, a median-joining haplotype network was reconstructed using NETWORK 4.2.0.1 []. We used the program LAMARC 2.02 [] to estimate migration rates (M) among populations in conjunction with each population's parameter θ (effective female population size multiplied by mutation rate). A Bayesian Metropolis Hastings Markov chain Monte Carlo search strategy was specified that consisted of a single final chain with 60 000 samples, a sampling interval of 1000, and a burnin of 6 000 000 steps. As MODELTEST 3.7 [] identified the HKY model [] as being most suitable for the COI data, we specified a transition/transversion ratio of 5.54 as determined in MEGA 4 [], as well as empirical base frequencies. Coalescent-based programs perform best when sample sizes are low [], so we limited our analyses to 20 individuals per population. However, in order to incorporate as much information as possible from the original data-sets of 35 individuals per population, we randomly sub-sampled each population three times to create different data-sets, each consisting of a different sample of 20 individuals per population. Values of θ and M, as well as their standard deviations and 95% confidence intervals, were then reported as the mean values of the three runs. Each run was repeated three times to check for consistency of results, and TRACER v. 1.4 was used to assess whether the program had run for sufficiently long, as indicated by effective sample sizes of no less than 200, and trendlines with stable likelihood values. A Mantel test in ARLEQUIN was used to test for isolation by distance (IBD). Most southern African coastal invertebrates with planktonic dispersal phases comprise mtDNA phylogroups whose distributions are limited by marine biogeographic disjunctions []. Pooling sequence data from different regional lineages will therefore incorrectly result in the detection of significant correlations between genetic and geographic distance [e.g., []], even in species whose phylogroups exhibit panmixia within their respective biogeographic region. As a minimum of three populations are required to test for IBD, we used only haplotypes of the subtropical mudprawn lineage. One matrix comprised genetic distances (ΦST values based on pairwise differences) and the other geographic distances (shortest long-shore distances among estuarine populations). To test for significant correlation between matrices, we specified 10 000 random permutations. […]

Pipeline specifications

Software tools Arlequin, LAMARC, ModelTest-NG, MEGA
Applications Phylogenetics, Population genetic analysis
Organisms Drosophila melanogaster, Caenorhabditis elegans