2.2 | RAD library construction and sequencing
The total DNA of each individual was extracted from the muscle tissue using the standard phenolchloroform method (Sambrook et al., 1989). DNA quality was first assessed by visually inspecting for signs of fragmentation on 1% agarose gels. To avoid the potential disproportionate representation of individuals in sequencing pools owing to the varying DNA quality across samples, we reextracted and reexamined any DNA samples showing degradation on agarose gels. Only nondegraded samples were processed further and quantified (twice) with a NanoDrop ND-1000 spectrophotometer. Each sample was then diluted to 10 ng/μL, requantified and pooled for each population. Next, each pooled sample was quantified (twice) with the spectrophotometer and equalized to 10 ng/μL to construct the RAD library.
To construct RAD libraries, DNA was fragmented by the restriction enzymeEco RI. A P1 adapter including a forward amplification primer, sequencing primer, and an 8-base pair (bp)-specific barcode was added to each of the fragmented DNA pools. The barcoded samples were then pooled and sheared randomly, and a P2 adapter was added to the sheared DNA fragments. DNA with a P1 adapter was selectively enriched by PCR amplification. Finally, DNA fragments of 300–500 bp were gel-purified and sequenced on the Illumina HiSeq2000 platform to generate 100-bp paired-end reads. All pools were sequenced on a paired-ends sequencing lane. The construction of RAD libraries and sequencing were carried out by BGI Tech Solutions Co., Ltd. (Shenzhen, China). Filtered reads for the RAD data pertaining to this study can be accessed through the NCBI GenBank Short Read Archive (PRJNA990519).