Computational protocol: Low Genetic Differentiation across Three Major Ocean Populations of the Whale Shark, Rhincodon typus

Similar protocols

Protocol publication

[…] Although a total of 68 whale shark samples were analyzed, some samples did not yield sufficient quantities of DNA to analyze all loci, and some DNAs did not yield useful data for all loci. Different numbers of samples are therefore reported for the different loci. In all cases, where fewer than 68 animals were used for analysis, this is indicated in the methodology. In individual tests for genetic differentiation (Structure and PCA), the single South African animal was treated individually. In analyses where animals were grouped into populations, this animal was included with the Indian Ocean group. Input files for various software were constructed using the program Create when possible . Locus data was initially checked for the presence of null alleles, stuttering and small allele dominance using the program MicroChecker 2.2.3 . Locus statistics and concordance with Hardy-Weinberg Equilibrium were calculated using FSTAT . GENEPOP 4.0 was used to test for linkage disequilibrium using the log likelihood ratio statistic (G-test), with the parameters, dememorization number = 10,000, number of batches = 1,000, number of iterations per batch = 10,000 . [...] A Bayesian approach using genotype data for individual animals was performed to detect any population structure across the entire data set using the program STRUCTURE 2.2 . STRUCTURE was run with assumptions of K = 1–5, using a burnin length of 50,000 and a run of 50,000 steps. All runs were repeated in triplicate at each K, and results were consistent across runs. Principle Components Analysis (PCA) is a method of detecting patterns of variation in complex data sets, and determining the extent to which individual patterns contribute to the variance of the data as a whole. PCA was conducted on individual multilocus genotypes using GenAlEx 6.1 with the standardized covariance method . All individuals that were genotyped at six or more loci (N = 43) were included, and the analysis was run without the Rtyp6 locus (see ).Traditional tests for population differentiation were performed by calculation of F-statistics using FSTAT and Microsatellite Analyzer (MSA) 4.05 . As numbers of animals from individual populations were small, animals were pooled into three same-ocean groups for analysis of population differentiation - Pacific, Caribbean and Indian. FSTAT was run without assuming Hardy-Weinberg equilibrium between populations, for 3,000 permutations. A matrix representing Nei's standard genetic distance (Ds) was produced using the program POPULATIONS 1.2.30 , . This analysis used animals scored for at least 4 loci (N = 59). [...] Calculation of effective population size (Ne) was performed using ARLEQUIN 2.0 . Data from all individual whale sharks was pooled into a single data set to obtain a global estimate. The number of mutations per generation, Theta (θH), was calculated from the expected homozygosity (HomE). Assuming that the population is in mutation-drift equilibrium, Theta(θH) is θH = (1−HomE)/HomE, where HomE = 1−HE, and HE is expected heterozygosity. To evaluate the utility of microsatellite genotypes as individual genetic tags, we estimated the probability of identity for individual loci and over all eight loci using Cervus 3.0.3 , . […]

Pipeline specifications