Genetic Population Structure of Black Drum in U.S. Waters

A main goal in fisheries science is to conduct robust scientific studies on fish species of interest in order to provide useful information that will support science-based management and conservation. One of the most important aspects of effective fishery management is identifying different population segments, which are often referred to as stocks (Waldman 1999). In management terms, “stock” usually refers to the segment of the population that is being exploited. However, defining the geographic boundaries of a stock can be particularly complex in marine species due to their capacity for long-distance dispersal and due to the difficulties in conducting direct field observations for many species (Avise 1998; Waples 1998). Many approaches can be used for stock analysis, but the most comprehensive picture of stock structure is obtained when data from a combination of study types (e.g., life history and genetics) are considered together (Jones and Quattro 1999; Waldman 1999). Most studies focus on economically important fish species whose population numbers are in decline or whose fisheries are under the threat of collapsing (Sinclair et al. 1985; Gold and Turner 2002; Kovach et al. 2010). However, more proactive measures should be taken to implement science-based management regulations before overfishing occurs, especially for species that may soon experience more intense pressure as fishers begin to depend more heavily on them due to the collapse and regional closure of other fisheries.

The Black Drum Pogonias cromis is the largest member of the family Sciaenidae and supports recreational and commercial fisheries throughout its range in U.S. waters (Sutter et al. 1986; Jones and Wells 2001). The Black Drum is an estuarine-dependent species found in the western Atlantic from the Bay of Fundy to Argentina and along the Gulf of Mexico (GOM) coast (Bleakney 1963; Sutter et al. 1986). Within the USA, Black Drum are currently managed as two separate stocks: (1) the GOM stock, managed by the Gulf States Marine Fisheries Commission; and (2) the Atlantic stock, managed by the Atlantic States Marine Fisheries Commission (ASMFC). A maximum age of 59 years has been reported for Black Drum along the U.S. Atlantic coast (Jones and Wells 1998), whereas a maximum age of only 43 years has been observed in the GOM (Beckman et al. 1990). Long-lived adults exhibit group-synchronous maturation of oocytes and batch spawning that occurs annually between sexual maturation and death (Pearson 1928; Fitzhugh et al. 1993). Earlier studies reported that Black Drum in the GOM reach sexual maturity by the end of their second year (Pearson 1928; Simmons and Breuer 1962; Sutter et al. 1986), but more recent investigations documented sexual maturation occurring much later. Murphy and Taylor (1989) reported that Black Drum in northeast Florida mature between 4 and 6 years; other studies in South Carolina (C. J. McDonough, unpublished data) and the GOM (Fitzhugh et al. 1993; Nieland and Wilson 1993) found similar results.

Black Drum form large aggregations and move to inshore or nearshore waters for spawning (in or near passes of bays and estuaries) during late winter into spring (Pearson 1928; Simmons and Breuer 1962; Jones and Wells 1998). In the GOM and northeast Florida, spawning occurs from January to May (Pearson 1928; Murphy and Taylor 1989; Nieland and Wilson 1993); at more northerly latitudes (e.g., Chesapeake Bay and Delaware Bay), spawning occurs later, from April to June (Richards 1973; Thomas and Smith 1973). Juvenile Black Drum recruit to shallow estuarine waters (e.g., tidal creeks) and move to deeper waters as they grow, until reaching sexual maturity and joining the spawning stock (Pearson 1928; Richards 1973). Adults move offshore in the late fall and exhibit the potential for long-range migration along the U.S. Atlantic coast, especially along the coasts of southeastern states (Jones and Wells 1998).

Results from tagging studies provide an estimation of the potential for long-distance movement in Black Drum. Most individuals in these studies (77–99%) were tagged and recaptured within the same state, but some moved across state lines (Murphy et al. 1998; ASMFC 2011). For example, Black Drum from tagging programs in South Carolina have been recaptured as far north as New Jersey and Delaware (ASMFC 2011), and two individuals that were tagged and released in Florida were caught a few months later in Chesapeake Bay (Murphy et al. 1998). Similarly, individuals tagged in Maryland have been recaptured as far south as Florida, demonstrating that movement occurs in both directions along the U.S. Atlantic coast (ASMFC 2011). Music and Pafford (1984) reported more limited movements of Black Drum tagged in Georgia, with fish being recaptured only as far south as Florida and as far north as North Carolina; the average time at large was 141 d, and tagged fish ranged in size from 150 to 414 mm TL. Within Texas waters of the GOM, Osburn and Matlock (1984) found little movement of Black Drum, even to adjacent bays. Currently, we do not have a rigorous understanding of when, how often, or how many Black Drum undergo long-distance migrations. More information from tagging studies will aid in determining the range and frequency of their movements and the potential mixing among populations, which relates to the genetic population structure of Black Drum.

Knowledge of the genetic population structure of Black Drum is limited. Gold et al. (1994) used mitochondrial DNA to examine genetic population structure of Black Drum collected in the GOM from 1987 to 1991. Those authors found that mitochondrial DNA haplotype frequencies did not significantly differ among samples from the GOM sites, but haplotype frequencies did suggest an isolation-by-distance pattern in which individuals from closer localities were more genetically similar than those from more distant sites. In another study, Gold and Richardson (1998a) examined variation in mitochondrial DNA among samples from the GOM and U.S. Atlantic for six of seven study species; although the 1987–1991 GOM data for Black Drum were included in the study, it was the one species for which no genetic data from the U.S. Atlantic were either presented or analyzed. In a discussion statement, Gold and Richardson (1998a) indicated that distinct subpopulations of Black Drum occurred in the GOM and western Atlantic based on unpublished data from 50 individuals sampled in Chesapeake Bay. However, no further details were provided concerning the unpublished data, such as when the samples were collected or methods used to obtain the genetic information. More extensive sampling of Black Drum, especially in the U.S. Atlantic, and the use of more modern genetic markers (e.g., microsatellites) are essential in providing scientifically rigorous tools to evaluate the genetic population structure of this species (1) between the U.S. Atlantic and GOM and (2) along the U.S. Atlantic coast. No studies have examined whether genetic population structure occurs for Black Drum along the U.S. Atlantic coast.

An understanding of genetic population structure is important for scientifically based management and conservation, including the accurate definition of geographic stock boundaries and biologically meaningful management units, which can be used for stock assessment purposes (Sinclair et al. 1985; Gold et al. 2001). For Black Drum, management regulations are established and monitored by individual states, but the potential movement of individuals across state jurisdiction lines may require greater cooperation among state agencies. The need for cooperative management was recently recognized by the ASMFC, and action was taken to develop and implement an Interstate Fishery Management Plan (ISFMP) in 2013 for Black Drum along the U.S. Atlantic coast. In addition to maintaining current management regulations, all Atlantic states were required (as of January 1, 2014) to establish a minimum size limit no less than 305 mm TL along with a maximum possession limit, although a specific maximum possession limit was not stipulated. Furthermore, all states must establish a minimum size limit no less than 356 mm TL by January 1, 2016 (ASMFC 2013). These new regulations increased the minimum size limit in Georgia (previously 254 mm TL) and created management requirements for Black Drum in North Carolina, which formerly had no regulations in place other than the reporting of commercial landings. However, management regulation differences still exist among states: Florida and South Carolina currently enforce maximum size limits of 610 and 686 mm TL, respectively (Florida also allows one fish over 610 mm TL for recreational anglers only), while the other states do not have maximum size limit regulations. In addition to establishing the new ISFMP, the ASMFC is conducting a coastwide stock assessment for Atlantic Black Drum. Genetic information will be important for this stock assessment and will assist in verifying appropriate management units and stock boundaries, which are necessary to develop and enforce scientifically based management regulations.

The main objective of the current study was to determine the genetic population structure of Black Drum in U.S. coastal waters. Using nuclear microsatellite markers, the stock structure of Black Drum between the U.S. Atlantic coast and GOM coast was re-evaluated. Additionally, small-scale spatial genetic structure along the U.S. Atlantic coast was examined. Better comprehension of the genetic population structure of this species in U.S. waters will help to inform future management decisions. Furthermore, an examination of the U.S. Atlantic coast at a detailed spatial scale may provide useful information for the Black Drum stock assessment and support for the new ISFMP.

METHODS

Sample collection.

Black Drum samples were collected from every state along the U.S. Atlantic coast from Florida to New Jersey and were collected from GOM waters of Texas, Mississippi, and Alabama (Figure 1). Fish were caught by various state agencies and other cooperative groups (Table 1) using a variety of gear types, including trawl, seine, gill net, pound net, trammel net, electrofisher, and hook and line. The location of capture and the TL of all individuals were recorded; total weight, sex, and age were also determined when possible. Fin clips were removed from sampled fish, preserved in 2.0-mL tubes containing 1.0 mL of a sarkosyl–urea solution (1% sarkosyl, 8-M urea, 20-mM sodium phosphate, and 1-mM EDTA), and stored at room temperature until analysis.

Table 1. Collecting agencies and the number of Black Drum collected from the waters of each state, including the total number of samples, number of adult samples, and number of juvenile samples (NJ = New Jersey; DE = Delaware; MD = Maryland; VA = Virginia; NC = North Carolina; SC = South Carolina; GA = Georgia; FL = Florida; AL = Alabama; MS = Mississippi; TX = Texas).

State	Collecting agency or agencies	Total samples	Adults	Juveniles
NJ	NJ Department of Environmental Protection, Division of Fish and Wildlife	8	0	8
DE	DE Department of Natural Resources and Environmental Control, Division of Fish and Wildlife	89	55	34
MD	MD Department of Natural Resources	1	1	0
VA	VA Marine Resources Commission	68	49	19
NC	NC Department of Environment and Natural Resources, Division of Marine Fisheries	101	1	100
SC	SC Department of Natural Resources, Marine Resources Division, Inshore Fisheries Research Section	314	7	307
GA	GA Department of Natural Resources	31	0	31
FL	FL Fish and Wildlife Conservation Commission	234	3	231
AL	AL Department of Conservation and Natural Resources, Marine Resources Division; and Fisheries Ecology Laboratory at Dauphin Island Sea Laboratory	54	40	14
MS	Gulf Coast Research Laboratory, University of Southern MS	151	1	150
TX	TX Parks and Wildlife	48	0	48

Age data were not available for the majority of specimens; therefore, 590-mm TL and larger individuals were classified as adults, and individuals smaller than 590 mm TL were classified as juveniles (based on Murphy and Taylor 1989; Table 1). Samples were collected from March 2012 to October 2013; additional samples from South Carolina were collected between April and August 2011. Because Black Drum generally do not move offshore until they reach sexual maturity, juveniles were sampled throughout the year. Adults in Delaware, Virginia, North Carolina, Florida, and Mississippi were only collected during the spawning season, thereby ensuring that they represented the spawning stock for the given region. Six of the seven adults captured in South Carolina were caught during July and August after the spawning season, whereas the other adult was collected in May at the end of the spawning season. Likewise, in Alabama, most samples were from adults caught during July; however, one adult was caught during June, and another was captured during October.

Isolation of DNA and microsatellite genotyping.

Isolation of DNA from the sarkosyl–urea solution containing Black Drum fin clips was performed using a Sprint Prep metal bead protocol (Seradyn, Indianapolis, Indiana). A 100-μL volume was removed from each sample vial and mixed in a round-bottom plate with 80 μL of 100% isopropanol and 10 μL of metal beads that bind DNA. The supernatant was drained using a magnetic plate, leaving the DNA bound to the metal beads. After five ethanol washes, the plate was allowed to dry, and 50 μL of tris-EDTA buffer (10-mM tris and 1-mM EDTA) were then added to elute the DNA from the metal beads. The supernatant was transferred to a clean plate for long-term storage in a −20°C freezer.

Because no microsatellite primers were available for Black Drum, confamilial species primers were used, as other studies of sciaenids and lutjanids have demonstrated the use of cross-species markers to be effective (Turner et al. 1998; Renshaw et al. 2007, 2009). Thirteen microsatellite primers for Spotted Seatrout Cynoscion nebulosus (Cneb07, Cneb23, Cneb25, Cneb31, Cneb37, Cneb41, Cne08A, Cne42C, Cne46C, Cne52C, Cne02D, Cne15D, and Cne612) and two primers for Red Drum Sciaenops ocellatus (Soc017 and Soc083) were screened for reliable amplification of Black Drum DNA (Turner et al. 1998; Chapman et al. 1999, 2002; Blandon et al. 2011; M. Tringali, Florida Fish and Wildlife Conservation Commission, personal communication). Markers were chosen based on the high levels of polymorphism observed in the Spotted Seatrout and Red Drum samples. The PCR protocol was optimized for proper annealing and extension temperatures by using temperature gradient protocols with a small set of samples (2–4) on an iCycler thermocycler (Bio-Rad Laboratories, Hercules, California). Amplified PCR products were visualized under ultraviolet light by using a Molecular Imager (Bio-Rad Laboratories) on a MetaPhor 3.5% agarose gel that was stained in a solution (<10%) of ethidium bromide. The final optimized reaction protocol consisted of an initial denaturation at 94°C for 2 min; 30 cycles of denaturation at 94°C for 30 s, annealing at 54°C for 1 min, and extension at 64°C for 1 min; and a final extension at 65°C for 60 min. Each reaction had a total volume of 11 μL and contained 1× HotMaster buffer (5 Prime, Inc., Gaithersburg, Maryland), 0.2 mM of each deoxynucleotide triphosphate, 2.5-mM MgCl₂, 0.2-mg/mL bovine serum albumin, 0.3 μM of forward and reverse primers, 0.03 units of HotMaster Taq DNA polymerase (5 Prime, Inc.), 2 μL of genomic DNA, and autoclaved milli-Q water.

Approximately 20 samples were run using the optimized PCR protocol to screen for polymorphic microsatellites that could be useful in evaluating genetic population structure. Six polymorphic microsatellite primers were identified, and forward primers were labeled with fluorescent dyes. Markers were arranged into two multiplex groups, and the ratios of primers were optimized for efficient PCR to save time and materials (Table 2). All samples were run with two negative controls for each multiplex group. After amplification, PCR products were separated on a CEQ 8000 automated sequencer (Beckman Coulter, Fullerton, California) with a labeled size standard. To ensure accuracy, genotypes were scored by two independent readers using CEQ 8000 Fragment Analysis Software (Beckman Coulter). Only samples genotyped at four or more loci were used in further analysis.

Table 2. Summary information for six microsatellite primers from Spotted Seatrout and Red Drum that were shown to amplify Black Drum DNA and were used to evaluate the genetic population structure of Black Drum in U.S. waters (FWC = Florida Fish and Wildlife Conservation Commission).

Multiplex	Locus	Repeat motif	Allelic size range (base pairs)	Number of alleles	Dye	Primer concentration (nM)	Source
1	Cneb23	(GA)14	118–150	12	D2	174.9	M. Tringali, FWC
	Cneb25	(TG)10(GT)5	125–157	12	D3	125.1	M. Tringali, FWC
2	Cneb31	(CA)8	90–124	11	D2	46.9	M. Tringali, FWC
	Soc083	(TG)19	120–150	10	D3	50.1	Turner et al. 1998
	Cne612	(GT)5n10(GT)11	122–156	16	D4	31.3	Chapman et al. 1999
	Cneb41	(TC)4(CT)3C(CT)7	162–186	13	D2	171.9	M. Tringali, FWC

Data analysis.

All loci were tested separately for each state to evaluate the effectiveness of the suite of six microsatellite markers across the U.S. range of Black Drum. GENEPOP version 4.0.10 (Raymond and Rousset 1995) was used to test for deviations from Hardy–Weinberg equilibrium and for linkage disequilibrium by using 10,000 dememorizations, 100 batches, and 5,000 iterations/batch. The frequency of null alleles at each locus was calculated using CERVUS version 3.0.3 (Kalinowski et al. 2007). Additionally, Micro-Checker version 2.2.3 (Van Oosterhout et al. 2004) was used to test for null alleles, large-allele dropout, and stuttering. A sequential Bonferroni correction was used to adjust significance levels (α) for all simultaneous analyses (Rice 1989). ARLEQUIN version 3.5.1.2 (Excoffier and Lischer 2010), FSTAT version 2.9.3.2 (Goudet 1995), and GENEPOP were used to calculate genetic diversity indices across all loci for Black Drum from each state. The indices included the number of alleles per locus (N_a), allelic richness (AR), expected heterozygosity (H_e; Nei 1987), observed heterozygosity (H_o), and inbreeding coefficient (F_IS).

Genetic population structure between U.S. Atlantic and GOM Black Drum and along the U.S. Atlantic coast was evaluated using pairwise comparisons of the genetic differentiation index F_ST (as calculated in ARLEQUIN) to determine the amount of genetic difference between populations. Typically, R_ST, which is an analog of F_ST based on differences in allele size rather than allele states, is used instead of F_ST for microsatellite data due to high mutation rates and typical stepwise changes in allele sizes. However, a test can be performed using the program SPAGeDi version 1.3d to determine whether allele-identity-based statistics perform better than allele-size-based statistics for a given data set (Hardy et al. 2003). Under the null hypothesis that differences in allele sizes do not contribute to population differences, the test used a randomization procedure in which the different allele sizes observed at each locus were randomly permuted among allelic states. Observed R_ST values were then compared with R_ST values calculated after permutation (pR_ST) to determine whether R_ST was significantly greater than pR_ST. For the current data set, R_ST was not significantly greater than pR_ST (P > 0.254), suggesting that allele size was not informative for population differences and that F_ST should be used rather than R_ST. Similar to pairwise comparisons of F_ST, exact G-tests comparing allelic frequency distributions were performed using GENEPOP with Markov chain parameters of 10,000 dememorizations, 100 batches, and 5,000 iterations/batch. An analysis of molecular variance (AMOVA) was also implemented in ARLEQUIN to determine the amount of variation both within and among various groupings of Atlantic states. Additionally, STRUCTURE version 2.3.3 (Pritchard et al. 2000) was used to evaluate Black Drum genetic population structure based on a clustering method that assigns individuals with similar genotypes to probable populations. The optimal number of populations or clusters (K) was determined by setting K equal to 1–9 (with 9 being the total number of states used in data analysis) and conducting three replicates for each K with a 100,000-iteration burn-in period and 100,000 Markov chain–Monte Carlo repetitions using sampling location as prior information. Results from STRUCTURE were then analyzed using the program STRUCTURE HARVESTER (Earl and von Holdt 2012) in which the appropriate K was determined according to Evanno et al. (2005), where L(K) is the estimated log probability of the data for a given K. Using CLUMPP version 1.1.2 (Jakobsson and Rosenberg 2007), membership coefficients for individuals were aligned across replicates for the chosen K to create a bar plot in Microsoft Excel. Additional analyses were run to characterize each population; these analyses included the calculation of H_e and F_IS using ARLEQUIN and GENEPOP and the estimation of effective population size (N_e) using LDNe version 1.31 (with the random mating model, lowest allele frequency of 0.02, and parametric confidence intervals [CIs]; Waples and Do 2008). Since multiple year-classes were evaluated, the N_e estimates represented some value between the effective number of breeders in one year-class and N_e. The N_e estimates were corrected with the bias-adjustment formula using adult life span and age at maturity from Waples et al. (2014). BOTTLENECK version 1.2.02 (Cornuet and Luikart 1996; Piry et al. 1999) was also used to test populations for recent bottlenecks by evaluating a population's heterozygosity excess compared with that expected at mutation–drift equilibrium. Interpretation of results was based on the Wilcoxon signed-rank test, as it is the most powerful and robust of the three tests when using fewer than 20 polymorphic loci. Lastly, because Gold and Richardson (1998a) reported an isolation-by-distance pattern in GOM Black Drum, a Mantel test for U.S. Atlantic Black Drum was performed in ARLEQUIN by using 10,000 permutations to test for a significant correlation between geographic and genetic distance matrices, which were created using coastline distances between sampling locations and pairwise F_ST values.

RESULTS

Genetic Population Structure: U.S. Atlantic and Gulf of Mexico Coasts

Exact G-tests of allelic frequency distributions showed significant genetic divergence between U.S. Atlantic and GOM populations of Black Drum; individuals from every Atlantic state were significantly different from those in every GOM state (df = 12, P < 0.001), with the exception of Georgia and Alabama after Bonferroni correction (Table 4). Pairwise comparisons of F_ST also indicated significant genetic divergence between Black Drum in the two regions. All comparisons produced F_ST values that were significantly different from zero (F_ST = 0.006–0.036, df = 1, P < 0.001) except for the Georgia–Alabama comparison and the Delaware–Alabama comparison (Table 4). When samples were pooled by region, the comparison between Black Drum in the U.S. Atlantic and those in the GOM produced a significant F_ST value of 0.014 (df = 1, P < 0.001).

Table 4. Results of exact G-tests of allelic frequency distributions (P-values, above the diagonal) and pairwise comparisons of F_ST values (below the diagonal) for Black Drum along the U.S. Atlantic coast and Gulf of Mexico coast (state abbreviations are defined in Table 1). Asterisks indicate values that were significant after Bonferroni correction (P < 0.0013).

State	DE	VA	NC	SC	GA	FL	AL	MS	TX
DE	—	0.018	0.273	0.066	0.717	0.005	<0.001*	<0.001*	<0.001*
VA	0.002	—	0.266	0.033	0.711	<0.001*	<0.001*	<0.001*	<0.001*
NC	0.001	0.000	—	0.263	0.287	0.032	<0.001*	<0.001*	<0.001*
SC	0.001	0.001	0.001	—	0.425	0.005	<0.001*	<0.001*	<0.001*
GA	−0.001	−0.001	0.007	0.004	—	0.753	0.017	<0.001*	<0.001*
FL	0.000	0.001	0.005*	0.003*	−0.002	—	<0.001*	<0.001*	<0.001*
AL	0.007	0.010*	0.012*	0.012*	0.003	0.006*	—	0.685	0.076
MS	0.013*	0.015*	0.016*	0.015*	0.009*	0.009*	−0.001	—	0.001*
TX	0.028*	0.028*	0.033*	0.036*	0.016*	0.025*	0.003	0.006*	—

The STRUCTURE results further supported the genetic difference between U.S. Atlantic and GOM Black Drum. The most appropriate grouping of the data comprised two populations (K = 2), as indicated by STRUCTURE HARVESTER results (mean L[K] = −19,312, SD = 5.57; ΔK = 56.5). Individuals from the U.S. Atlantic were largely assigned to their own population; the highest proportion of GOM membership was assigned to samples from Florida, with membership proportions ranging from 0.022 to 0.289 but averaging only 0.041. Individuals from the GOM were also largely assigned to their own population, but comparatively they had a much higher proportion of Atlantic membership, which ranged from 0.201 to 0.758 with a mean of 0.541 (Figure 2). The U.S. Atlantic and GOM populations both showed moderate levels of genetic diversity (H_e = 0.683 and 0.740, respectively) and small F_IS values (0.013 and 0.024, respectively). The N_e for the U.S. Atlantic population was a negative estimate (95% CI = 1,992.9 to infinity), which can be interpreted as an estimate of infinity for N_e; the N_e of the GOM population was estimated at 1,172.5 (95% CI = 400.9 to infinity). Neither population showed evidence of recent bottlenecks, as the Wilcoxon signed-rank tests for heterozygosity excess were not significant under the two-phase mutation model (U.S. Atlantic: P = 0.976; GOM: P = 0.656) or the stepwise mutation model (U.S. Atlantic: P = 1.000; GOM: P = 0.992).

In the GOM, Black Drum sampled in Mississippi and Texas waters were significantly different from each other according to pairwise F_ST comparisons (F_ST = 0.006, df = 1, P < 0.001) and exact G-tests of allelic frequency distributions (df = 12, P < 0.001). Individuals from Alabama were not significantly different from those in either Mississippi or Texas (Table 4).

DISCUSSION

The genetic population structure of Black Drum in U.S. waters has not been examined for over 20 years; with the recent development of the ISFMP for Atlantic Black Drum and a stock assessment currently in progress, new genetic data have come at an opportune time. Not only is the current study the only one to evaluate genetic population structure for Black Drum since Gold et al. (1994), but it is also the first published study to (1) examine Black Drum genetic population structure between the U.S. Atlantic coast and GOM coast as well as among areas along the U.S. Atlantic coast; and (2) investigate this species’ genetic population structure by using nuclear microsatellite markers. Our finding that the suite of six microsatellite markers is valid for use with Black Drum adds to the growing number of genetic studies that have verified the use of confamilial species markers (Turner et al. 1998; Renshaw et al. 2007, 2009).

Significant genetic divergence between Black Drum along the U.S. Atlantic coast and those in the GOM supports the currently managed stock structure of Black Drum in U.S. waters. Although not every comparison between individuals from U.S. Atlantic states and GOM states was significant, the overall pattern was clear and supported by exact G-tests of allelic frequency distributions, pairwise comparisons of F_ST, and results from STRUCTURE analysis. Failure to detect significant differences in every comparison of Black Drum from Atlantic states versus GOM states highlights the importance of sampling multiple locations within a region for studies examining genetic population structure. Studies have also reported differences in life history aspects of Black Drum in the U.S. Atlantic and GOM, such as slower growth, smaller size at age, and smaller maximum size of fish in the GOM (Beckman et al. 1990; Jones and Wells 1998). Life history differences could be the result of environmental differences between the two regions, but they could also be genetically based. Because neutral microsatellite markers were used in the current study, it cannot be determined whether these life history differences are driven by selection, but the differences suggest that genetic divergence between the two regions has biological significance. Genetic divergence between individuals in the U.S. Atlantic and those in the GOM has been documented for other marine species, including the Red Drum, Spotted Seatrout, Black Sea Bass Centropristis striata, and Greater Amberjack Seriola dumerili (Avise 1992; Gold and Richardson 1998a, 1998b). Multiple explanations have been proposed for the common patterns of regional divergence among species, including historical separation due to climatic changes during Pleistocene glaciation (Avise 1992), along with present-day factors, such as an absence of suitable habitat in the area separating the two regions (e.g., narrow continental shelf along the southeastern coast of Florida) or ocean current patterns that limit movement between the Atlantic and the GOM (Gold and Richardson 1998a).

As has been speculated for other marine species, Black Drum in the U.S. Atlantic and GOM may have been separated historically, but results from the current study illustrate a pattern of recent or current gene flow between the two regions. All but two pairwise comparisons of F_ST between samples from U.S. Atlantic states and GOM states were significant; however, the F_ST values were small, including the F_ST from the comparison between the pooled samples representing each region. The small F_ST values suggest that some degree of gene flow is occurring or has occurred between the two regions, thus preventing stronger divergence (i.e., larger F_ST values). The STRUCTURE results also indicated gene flow between Black Drum in the U.S. Atlantic and those in the GOM. However, the proportion of Atlantic membership assigned to GOM individuals was higher than the proportion of GOM membership assigned to Atlantic individuals—the opposite of what we would predict in terms of ocean current patterns influencing movement between the two regions. Current velocities are strongly eastward from the Florida Keys through the Florida Straits and northward along the east coast of Florida, which should result in greater movement of individuals from the GOM into the Atlantic than vice versa (Gold and Richardson 1998a). Adult Black Drum are strong swimmers that can potentially swim against the strong ocean currents around the Florida peninsula, but this possibility cannot be evaluated without tagging studies to provide data on movement between the Atlantic and the GOM.

The large proportion of mixed ancestry in GOM individuals may also be the result of gene flow with Central American and South American Black Drum, as the species’ range extends down the western Atlantic coast to Argentina. There is a lack of published research on Black Drum populations from Central America and South America; a single study from Argentina documented reproductive activity, providing evidence for a spawning season from October to January (Macchi et al. 2002). No tagging studies have yet been published to determine how far Black Drum travel in the Central American and South American range or whether mixing occurs among individuals from South America, Central America, and North America. The GOM may serve as a mixing zone for individuals along the western Atlantic coast, which could explain the lack of complete GOM ancestry in the GOM individuals sampled in the current study. Additional samples of Black Drum from other GOM locations (especially along the GOM coast of Florida) and from Central America and South America could help to clarify the pattern of gene flow between the Atlantic and GOM regions and other potential contributing factors.

Along the U.S. Atlantic coast, weak but significant genetic divergence between Black Drum from our sampling locations was identified. The current results suggest that this small-scale genetic structure only occurs along the southeast, specifically from North Carolina to Florida, and is primarily due to differences between Black Drum from the Tequesta, Florida location and the Carolinas. However, the F_ST values for the comparisons between individuals from the Carolinas and Florida were small, and they were lower than the F_ST values for comparisons of the U.S. Atlantic and GOM. The small amount of variation among the groups as determined by AMOVA also suggested weak structure. An isolation-by-distance pattern was observed along the U.S. Atlantic coast but only from North Carolina to Florida. Based on the previous observation of an isolation-by-distance pattern for Black Drum in the GOM (Gold et al. 1994), along with findings from tagging studies, we predicted that Black Drum would follow an isolation-by-distance pattern along some portion of the U.S. Atlantic coast. Black Drum are capable of extensive offshore migrations and have been documented to travel up to 1,370 km in just a few months (Murphy et al. 1998), but tagging studies have reported that most individuals are recaptured either from the same state in which they were released or from adjacent states (Music and Pafford 1984; ASMFC 2011). The dependence of Black Drum on estuaries, especially for juvenile habitat (Pearson 1928; Sutter et al. 1986), also supports the isolation-by-distance pattern found in the current study, although no published studies have investigated natal homing in Black Drum. Red Drum and Spotted Seatrout have similar estuarine-dependent life history strategies (Pearson 1928) and also exhibit isolation-by-distance patterns in the GOM (Gold and Richardson 1998a). Additionally, O'Donnell et al. (2014) provided evidence for an isolation-by-distance pattern for Spotted Seatrout along the southeastern U.S. Atlantic coast from North Carolina to Georgia.

Genetic differences in Black Drum along the U.S. Atlantic coast are not solely determined by geographic separation. The isolation-by-distance pattern was not significant when samples from Delaware and Virginia were included, and Black Drum from those states showed no significant differences relative to individuals in the other Atlantic states. In this case, the lack of genetic structure observed between Black Drum from the northern and southern Atlantic states may be influenced by the life history patterns of the species. Adults are primarily caught in their northern range, with studies noting a scarcity of young Black Drum in Chesapeake Bay and Delaware Bay. Jones and Wells (1998) reported an absence of 1- to 5-year-old Black Drum in their samples, which could be partly an artifact of the gear used in this region—mainly gill nets with larger (33-cm) stretch mesh (used by commercial fishermen) and hook and line (used by recreational anglers). However, Jones and Wells (1998) indicated that fishermen had not seen small Black Drum in smaller pound nets and gill nets with 7.6–15.2-cm stretch mesh. Additionally, Frisbie (1961) presented records of the few young Black Drum caught in Chesapeake Bay and Delaware Bay from 1912 to 1960, and no fish over 267 mm TL were recorded. Lastly, Richards (1973) reported that Black Drum between 220 and 800 mm TL were not readily available in Virginia waters. Most of the Delaware and Virginia samples collected in the current study were from adults (≥590 mm TL), and there was an apparent gap in the sizes of the juveniles that were caught. The gap in sizes for Virginia samples (no individuals between 281 and 441 mm TL were caught) was smaller than that for Delaware samples (no fish between 253 and 704 mm TL were collected). The majority of the juveniles that were caught in Virginia and Delaware waters were smaller than 355 mm, which is the average observed TL for 1-year-old Black Drum (Murphy and Taylor 1989). A few juveniles that were collected in Virginia exceeded 355 mm TL, but most were close to 590 mm and thus were approaching sexual maturity. Therefore, the majority of juveniles sampled from Virginia and Delaware were 1 year old or younger based on their length. The absence of juvenile Black Drum after their first year and the lack of genetic structure observed between northern and southern Atlantic states in the current study may be partially explained by patterns of Black Drum migration to and from Chesapeake Bay and Delaware Bay.

Previous studies have proposed that Black Drum undergo a seasonal migration along the U.S. Atlantic coast, moving southward and offshore during the fall and moving northward and inshore during the spring (Richards 1973; Jones and Wells 1998). Only larger and older fish return to Chesapeake Bay and Delaware Bay in the spring for spawning, which may indicate that juveniles recruit to the estuaries of these bays but only remain in these areas until they reach a large enough size during their first year to move offshore and migrate south in the fall, taking up residency in southern estuaries and mixing with juvenile recruits from southern Atlantic states (Frisbie 1961; Jones and Wells 1998). Since studies in Chesapeake Bay and Delaware Bay have documented an absence of younger and smaller Black Drum after their first year, juveniles originating from the bays may remain in southern estuaries until they reach sexual maturity between 4 and 6 years and then make the migration northward during the spring to spawn in the bays (Figure 3). In the current study, Black Drum sampled from North Carolina to Florida were mainly juveniles, including fish in the size ranges missing from Delaware and Virginia; thus, we hypothesize that some of these juveniles may be recruits from Chesapeake Bay and Delaware Bay. Murphy et al. (1998) tagged and released two individuals in the Indian River Lagoon system along the Atlantic coast of Florida in February, and those fish were recaptured in Chesapeake Bay during late May and early June. Results from the Murphy et al. (1998) study demonstrate that adult Black Drum can travel long distances along the U.S. Atlantic coast. Some may even do so during the spawning season, possibly spawning along the Atlantic coast as they move northward since they are capable of spawning every 3–4 d (Fitzhugh et al. 1993; Nieland and Wilson 1993). Black Drum adults are still found in Florida during the spawning season (Murphy and Taylor 1989), so apparently not all fish make the long-distance migration. Overall, the southward migration of juveniles from Chesapeake Bay and Delaware Bay—with their possible return as sexually mature adults in the spring years later—offers a potential explanation for the genetic patterns seen in the current study. If, after they reach a certain size, juveniles from Chesapeake Bay and Delaware Bay mix with juveniles in the estuaries of southern Atlantic states and then return to the bays as adults, we would predict no significant genetic divergence between the samples collected from Delaware and Virginia and samples from all other Atlantic states. Samples of juveniles collected in North Carolina, South Carolina, Georgia, and Florida may include not only recruits from these states but also recruits from Delaware and Virginia. More research is needed to explore this possible explanation for our findings.

The observed lack of genetic structure between Black Drum from northern Atlantic states and those from southern Atlantic states may also be influenced by current patterns along the U.S. Atlantic coast. The Gulf Stream current may act as a mixing force for individuals between southern and northern regions, creating passive movement of individuals from the south to the north. However, we could find no data from other studies to support this hypothesis. Instead, it has been suggested that the divergence of the Gulf Stream current off the U.S. Atlantic coast at Cape Hatteras, North Carolina may lead to differences between northern and southern marine populations due to environmental differences (i.e., temperature). Cape Hatteras is considered an important zoogeographic boundary (Briggs 1974). For Summer Flounder Paralichthys dentatus, Wilk et al. (1980) documented meristic and morphological differences between northern and southern populations, whereas Jones and Quattro (1999) failed to find any significant genetic population structure based on the Cape Hatteras boundary. Genetic studies of Weakfish Cynoscion regalis, Bluefish Pomatomus saltatrix, and sea catfishes (Ariidae) failed to find significant genetic divergence between the northern and southern populations as well (Avise et al. 1987; Graves et al. 1992a, 1992b). Significant genetic divergence centered around Cape Hatteras was documented in the Oyster Toadfish Opsanus tau; it is a sedentary species that produces demersal eggs and larvae, a life history strategy leading to decreased gene flow between regions, in stark contrast to the above-mentioned species and Black Drum (Avise et al. 1987). The current study appears to be in agreement with other genetic studies of mobile marine species in that we documented a lack of genetic divergence between northern and southern populations along the U.S. Atlantic coast.

Overall, the results of the current study support the continued management of Black Drum in U.S. waters as two separate stocks: Atlantic and GOM. There is little concern about the genetic health of these populations at this time, as they exhibit moderate genetic diversity and no sign of inbreeding. Both populations appear to have large N_e values and no indication of recent population bottlenecks. The present results provide important baseline information that can be used for future monitoring of both populations. The current study also supports the management of Black Drum along the U.S. Atlantic coast as one unified stock. Although significant genetic differences were observed among Black Drum from the southern Atlantic states, those differences were weak. The larger pattern—namely a lack of genetic divergence between individuals from northern and southern Atlantic states, indicating the potential mixing of Chesapeake Bay and Delaware Bay individuals with southern populations—necessitates the need for common management regulations along the entire U.S. Atlantic coast. The development of the ISFMP by the ASMFC in 2013 was the first step toward the cooperative management of Black Drum. Thus far, the new management requirements have only affected two Atlantic states: Georgia and North Carolina. Since most Black Drum do not reach sexual maturity until attaining a size close to 590 mm TL, the recreational fishery primarily affects immature subadults, which could have an impact on the spawning stock biomass of the population, particularly in states where the fishery is mostly estuarine based. Maximum size limits (currently in place only in Florida and South Carolina at 610 and 686 mm TL, respectively) could help to ameliorate this impact and would assist in protecting the spawning adult stock. No published data exist to suggest that Black Drum are currently in danger of overfishing, but the ongoing stock assessment will both provide more information on the fishery status and help to determine whether overfishing is an impending possibility. Future studies should focus on combining information from different biological perspectives (e.g., genetics, tagging, and life history) to build a comprehensive picture of Black Drum stock structure that can help to better define stock boundaries and management units and ultimately provide scientifically based management and conservation of this species.