Refer to the following publication for complete methodology details:
Goetze, E., Hüdepohl, P., Chang, C., Iacchei, M., Van Woudenberg, L., Peijnenburg, K. T. C. A. (2016) Ecological dispersal barrier across the equatorial Atlantic in a migratory planktonic copepod. Progress in Oceanography – AMT special issue. doi: 10.1016/j.pocean.2016.07.001

In summary (excerpted from above):
Bulk plankton samples were collected on Atlantic Meridional Transect Cruise 22 (AMT22) between 10/13/2012 and 11/19/2012. Oblique tows were conducted with bongo nets (200 um, 333 um), towed between on average 324 m depth and the sea surface. A General Oceanics flowmeter (2030RC) mounted in the mouth of the 200 um net was used to measure seawater filtered during the tow. Plankton from the 200 um mesh net was bulk preserved immediately in 100% ethyl alcohol, the alcohol was changed to fresh within 12–24 h of collection, and samples were stored at -20 C. Plankton from the 333 um mesh net was sorted live at sea, and Pleuromamma xiphias specimens were preserved immediately in RNALater (Ambion), followed by cryopreservation in liquid nitrogen, and long-term storage at -80 C.

Specimens included in the genetic analyses in this study were collected at 18 stations, located between 39 38.82N and 40 4.39S latitude. The majority of genetic analyses focused on stations with sufficient sample size for population-level inference (N > 42), including AMT22-09 through AMT22-29 and AMT22-45 through AMT22-68. Specimens used in genetic analyses were primarily RNALater-preserved, but some specimens were included from ethanol-preserved samples to achieve the minimum target sample size of 45 individuals per station. When a major genetic break across the equatorial region was identified, we included samples from stations AMT22-31 through AMT22-43 to assess the genetic composition of populations across this region. Population-level analyses were not conducted on these latter stations due to small sample sizes and low abundance in this region, but sequences from these animals were included in the haplotype network. DNA was extracted from individual P. xiphias adults using the DNeasy Blood & Tissue kit (Qiagen), following the manufacturer’s protocol, with the exception of longer elution incubation times (Goetze, 2011). The second of two elutions for each individual was used in this study. Polymerase Chain Reaction (PCR) amplification of a 681-bp fragment of mtCOI was conducted with primers and PCR and sequencing protocols as described in (Goetze, 2011). Forward and reverse sequences from each individual were aligned and checked for errors in Geneious (v7.1.8, Biomatters). Consensus sequences for all individuals were aligned using MUSCLE (Edgar, 2004), and unique mtCOI haplotypes were identified using FABox (http://users-birc.au.dk/biopv/php/fabox/). MtCOI sequences representing unique haplotypes are available under GenBank accession numbers KT429028–KT429159. A minimum spanning haplotype network was inferred for all mtCOI sequences using Population Analysis with Reticulate Trees (PopART; http://popart.otago.ac.nz), in order to investigate geographic patterns in the distribution of haplotypes across the Atlantic.

Population subdivision across samples was investigated using several approaches. Refer to publication cited above.