1 | INTRODUCTION

Gene flow from planktonic larval dispersal has the potential to maintain genetic panmixia over large distances in the marine realm (Shaklee & Bentzen, 1998), although this potential is not always realized. Possible causes of departures from panmixia include historical biogeographic separation, physical oceanographic barriers, larval behavior, sweepstakes reproductive success (SRS), and natural selection (reviewed in Hellberg, 2009). Disentangling these causes is challenging because they can act synergistically. Thus, a genetic break formed by a vicariant event could subsequently be maintained by selection (Schneider-Broussard, Felder, Chlan, & Neigel, 1998), or larval behavior could exploit oceanographic features to limit dispersal (Kingsford et al., 2002). Sometimes it is possible to exclude mechanisms a priori, thus in species with low fecundity SRS can be eliminated as the cause of chaotic patchiness (Cornwell, Fisher, Morgan, & Neigel, 2016). More generally, distinctive patterns of genetic population structure are considered signatures of specific mechanisms (Hellberg, Burton, Neigel, & Palumbi, 2002). Extensive sampling is needed to adequately characterize these patterns, which can appear at different scales across time, space and life stage. Furthermore, numerous marker loci are needed to detect and quantify the typically low levels of population structure in marine populations (Waples, 1998). With the above considerations in mind, we undertook a survey of single nucleotide polymorphisms (SNPs) in the blue crab,Callinectes sapidus , an economically important, well-studied species (Kennedy & Cronin, 2007).
The life cycle of the blue crab favors dispersal away from natal habitats (Epifanio, 2007). Females release planktonic larvae offshore where they are transported by currents for 4-6 weeks. Megalopae (post-larvae) migrate to coastal brackish water habitats where they settle and metamorphose into juveniles. After mating, mature females migrate to offshore spawning sites, sometimes travelling hundreds of kilometers (Aguilar et al., 2005; Gelpi, Fry, Condrey, Fleeger, & Dubois, 2013). Most surveys of genetic variation in North American blue crabs have found no detectable genetic differentiation (Berthelemy-Okazaki & Okazaki, 1997; Lacerda et al., 2016; McMillen-Jackson & Bert, 2004; Yednock & Neigel, 2014) or very little (McMillen-Jackson, Bert, & Steele, 1994; Plough, 2017) over distances up to thousands of kilometers. In the surveys that found slight differentiation, it was detected near the limits of their statistical power (discussed in Yednock & Neigel, 2014), which suggests that additional genetic structure is present at undetectable levels. A few studies have found slight temporal or life-stage shifts in the genetic composition of blue crab populations (Feng, Williams, & Place, 2017; McMillen-Jackson et al., 1994; Yednock & Neigel, 2014). Temporal shifts can cause chaotic genetic patchiness, a pattern of fluctuating small-scale spatial variation first described in limpets (Johnson & Black, 1982, 1984). Kordos and Burton (1993) reported anomalously high levels of temporal and spatial differentiation for allozyme loci in blue crabs from the coast of Texas. It is possible that pronounced genetic structure such as they described develops episodically or at specific locations in blue crab populations, but see Sullivan and Neigel (2017) on the misidentification of specimens as a plausible explanation for some of these anomalous findings.
The native range of C. sapidus is disjunct along the Atlantic coast of the Americas and encompasses at least two distinct genetic units. The northern portion of the range extends from Massachusetts in the U.S. to Venezuela while the southern portion extends from Bahia, Brazil to Mar del Plata in Argentina (Santos & D’Incao, 2004). Yednock and Neigel (2014) found that blue crabs from the northern Gulf of Mexico were genetically different from those from Venezuela (FST for each locus between 0.06 and 0.68) but not from Mexico. Rodrigues and co-workers (2017) found two mitochondrial lineages in blue crabs, one in both portions of the species’ range and the other exclusive to the southern portion. It is likely that unintended sampling of multiple species ofCallinectes has been a source of error in blue crab ecological and population genetic research, especially for studies that included early life stages that are difficult to identify to species (Sullivan and Neigel, 2017).
Our sampling program was designed to be sensitive to both low levels of genetic differentiation and episodic genetic structure. We sampled individuals from multiple locations, time points, and life stages to characterize patterns of genetic differentiation rather than simply test for its presence. Some SNPs were in genes targeted for their presumed functions. These criteria led us to a genotyping method, the Infinium Assay (Illumina), which is seldom used for population genetics of non-model organisms. We were able to use the Infinium assay because an annotated transcriptome allowed us to identify SNPs in protein-coding regions and a draft genome helped us locate and avoid introns.
Our findings support the main conclusions of most previous studies: we detected little genetic population structure in blue crabs from the Gulf of Mexico and North Atlantic. In addition, there was no evidence of SRS: we found neither chaotic genetic patchiness or closely related individuals. Overall, it appears that levels of gene flow are high among North American blue crabs as expected from their biology. Although these findings might appear unremarkable, they point to an opportunity to observe the effects of natural selection on genetic structure against a background that is relatively free of other structuring mechanisms. Here we focus on detection of genetic differentiation in blue crabs and its plausible causes, exclusive of natural selection.