Computational protocol: Subtle genetic structure reveals restricted connectivity among populations of a coral reef fish inhabiting remote atolls

Similar protocols

Protocol publication

[…] DNA for genotyping was extracted with the high-throughput membrane-based DNA extraction protocol of Quality and quantity of genomic DNA was ascertained through gel electrophoresis using 1% standard agarose (Amresco, Ohio, USA), and then diluted by one-third with millepore purified water (final concentration ca 10–20 ng) before PCR. The development of the microsatellite library, characterization of final 10 loci, and genotyping procedure were described in . To mitigate and report scoring error of microsatellites, quality control procedures suggested by and were implemented. Specifically, genotyping each individual involved the implementation of negative controls and the visual inspection of all automated allele calls, and individuals with suspect electropherograms were repeated. A genotype error rate (0.83%) was measured by repeating the genotyping procedure, from DNA extraction through to final allele scoring, using a subset of blind samples (n= 24) selected from three sites randomly spread across the sampling area.Allelic patterns of the 580 adult fish collected from 13 sites, and 98 recruits collected from four sites, were calculated with GenAlEx v6.3 (). The number of alleles (NA), the unbiased expected heterozygosity (HE), the fixation index (FIS) at each of 10 microsatellite loci, and the number of private alleles at each site averaged across loci are presented in and . Tests for Hardy–Weinberg and linkage disequilibrium were conducted with FSTAT v2.9.3 () using the inbreeding coefficient FIS and significance levels were based on 1000 permutations of alleles among individuals within sites and were adjusted with sequential Bonferroni correction for multiple tests when P < 0.05. Micro-Checker v2.2 () was used to detect and adjust for null alleles. Because, the vast majority of samples amplified across all loci (i.e., no obvious null homozygotes), we used Brookfield equation 1 to estimate null allele frequencies. [...] To infer the strength of genetic connectivity across a hierarchy of spatial scales among the atolls of north-west Australia, we measured the amount of genetic variation that was geographically structured among adult samples with an analysis of molecular variance (AMOVA) framework in GenAlEx v6.3 (). We partitioned genetic variation between atoll systems (FRT), among sites relative to variation within each atoll systems (FSR), and among sites relative to overall variation (FST). Additionally, we calculated the variation partitioned among sites at Rowley Shoals and Scott Reef system relative to the variation within that particular system (FSR Rowleys and FSR Scott). To account not only for the high degree of variation within populations of microsatellite markers, but also for the effects that potential differences in effective population sizes might have on subdivision, we also calculated a standardized measure of all the F-statistics (F′RT, F′SR, and F′ST) according to the method of . To visualize the genetic relationships among adult samples, we performed a Principal Coordinates Analysis (PCoA; sensu ) with pairwise DS () and FST estimates calculated in GenAlEx v6.3 (). DS performed well in studies that evaluated the effectiveness of different genetic distances (; ), and provides a complimentary and independent comparison to the commonly used pairwise FST. Pairwise matrices of FST (along with F′ST) and DS estimates are given in and . We assessed the significance of spatial differentiation among allele frequencies of within each hierarchical grouping used in the AMOVA with a Fisher exact test implemented in Genepop v4.0 (). For this powerful test that is well suited to unbalanced samples sizes (), we used Markov chain parameters of 1000 iterations, 1000 batches, and dememorization number of 100.To investigate further patterns of genetic connectivity and changes in genetic composition across generations, we assessed the genetic relationships among adults and recruit samples. Specifically, we tested for significant differentiation between recruit and adult samples at the four sites in which recruits and adults were collected (i.e., RS3, SL3, SL1, and SS2) with the exact tests implemented Genepop v4.0 (using the same Markov Chain parameters as spatial tests). Furthermore, to test explicitly for the effects sweepstakes reproduction, we calculated allelic richness (RS), the degree of relatedness within samples (Rel; ), heterozygote deficiency (FIS:), and levels of geographic structure (FST; ), and tested for significant differences with randomization procedures (10,000 permutations) with FSTAT v2.9.3 (). To visualize the temporal genetic relationships among recruit and adult samples relative to the amount of overall spatial variation, we conducted a PCoA in GenAlEx of pairwise DS and FST estimates among all adult and recruit samples. Finally, because differences in levels of subdivision were detected among adult samples collected from the Rowley Shoals and Scott Reef systems (see Results) suggested differences in intensity of sweepstakes reproduction at each system, we tested also differences in allelic richness (RS) and heterozygote deficiency (FIS; ) between adult samples from each atoll system with randomization procedures (10,000 permutations) as implemented in FSTAT v2.9.3 ().In addition to the above genetic analyses, we explored spatial patterns of genetic structure and propensity for self-recruitment of C. margaritifer with a number of methods. However, because these analyses yielded little additional information, we provide a brief description of methods and results in the Supporting Information. […]

Pipeline specifications

Software tools GenAlEx, Genepop
Application Population genetic analysis
Diseases Genetic Diseases, Inborn