Perna perna metadata and genotypic data

Studies investigating gene flow in sessile or sedentary marine species typically draw conclusions about larval dispersal by investigating genetic structure of adults. Here, we generated microsatellite data from adults, recruits, settlers and planktonic larvae of the brown mussel, Perna perna, from the south-east coast of South Africa, and identified a consistent mismatch in genetic structure between the adults and all earlier life stages. While adults could be assigned to two major geographical groups (western and eastern), most of the early-stage mussels were strongly affiliated with the eastern group. This result suggests that few of the early-stage individuals present in the western portion of the sampling range will become part of the adult population. Our findings highlight the importance of post-settlement processes as key drivers of population structure and caution against the exclusive use of genetic data generated from adults to assess population connectivity facilitated by the dispersal of planktonic propagules.

Methods
Sample collection and data generation
Adult, recruit, settler and larval specimens of Perna perna were sampled along the western edge of the region where the ranges of the species’ two evolutionary lineages overlap. Fieldwork was conducted between March 2016 and February 2017. The total number of samples collected per life-history stage was 344 adults from eight sites, 86 recruits and 99 settlers collected from four of these sites, and 78 larvae from five adjacent nearshore sites. Larvae were collected from the surface waters between 900 and 2400 m from the shore using a submersible 2.2 KC Denmark 23.580 plankton net during two consecutive spawning periods.

Genomic DNA from adults, recruits and settlers was extracted using the CTAB protocol (Doyle and Doyle 1987; Doyle 1991). Larval specimens were extracted using a modified one-step incubation/denaturation protocol (Badenhorst and Roodt-Wilding 2007). Samples were genotyped using four polymorphic microsatellite loci designed by Coelho et al. (Coelho et al. 2012), with a fluorescently labelled M13 primer incorporated into the PCR product (Schuelke 2000). For each specimen, PCR products were pseudo-multiplexed (pooled) and genotyped on an ABI 3130xl Genetic Analyzer (Applied Biosystems). Geneious was used for peak calling, bin alleles and produce a table of genotypes. Since this is diploid data, there are two alleles per locus, coded in separate columns. Missing data is depicted with 0 (zero). Tests for genotyping errors, null alleles and allelic dropouts were performed in MICRO-CHECKER v 2.2.3 (Van Oosterhout et al. 2004).