Gene flow between subpopulations of gray snapper (Lutjanus griseus) from the Caribbean and Gulf of Mexico

Introduction

Establishing effective fisheries regulations is a complex and multidisciplinary task (INAPESCA, 2012). In recent decades, it has been determined that delineating stock boundaries and knowing the level of genetic connectivity among stocks is useful for establishing appropriate management and conservation strategies (Villegas Sánchez et al., 2014). Improper management can lead to genetic diversity loss and increased inbreeding within genetically isolated populations with negative effects for the survival of the populations (Urbiola-Rangel & Chassin-Noria, 2013; Villegas Sánchez et al., 2014). With the use of genetic markers, diversity and the level of genetic connectivity can be estimated between populations at different geographic scales.

The Gray snapper (Lutjanus griseus) is an economic and ecologically important fishery species and can be highly abundant throughout its range, yet there is a paucity of information for its fishery management (Lindeman et al., 2016). Gray snapper fisheries are subjected to fishery regulations in some countries, for example, in the Everglades National Park, United States, a catch limit of 10 individuals per person was established in the 1970’s (Claro & Lindeman, 2008). In Cuba the fishery is closed in June, during the breeding season (Claro & Lindeman, 2008). Nevertheless, in Mexico, although it’s being captured by fishers, it has not been classified as overexploited or subjected to fishery overexploitation, thus no closed season or other regulation has been established (INAPESCA, 2018); this lack of regulation could place populations at risk in the future (Costello et al., 2012).

The gray snapper presents a wide distribution from North Carolina, United States to southern Brazil, in the countries where it is found it represents an important reef fishery resource because of its high quality meat and market value (Claro & Lindeman, 2008). The gray snapper is a predator that feeds on a wide variety of organisms in different habitats including estuaries, mangroves and seagrass beds, whereof changes in their populations have great impacts in other elements of the community (Claro & Lindeman, 2008; Rocha & Molina, 2008). Migratory movements of adult individuals are mainly local and sexual maturity occurs when they are 1 year old reaching a total length between 260 and 280 mm. The pelagic larval period of this species is 25–26 days and larval settlement has been recorded between 30 and 40 days after hatching (Claro & Lindeman, 2008). Its large geographic range and month-long larval duration indicate that connectivity among populations may be widespread.

Despite its importance as a fisheries species, there is relatively little knowledge about the stock structure of gray snapper, however stocks appear to be declining in some places. Commercial landings in the US South Atlantic region (North Carolina to Florida) have been declining since the 1950’s and populations in the Florida Keys are thought to be potentially over-fished (Ault, Bohnsack & Meester, 1998). In Cuba, gray snapper is abundant; however, fisheries landings have declined over the past 30 years as have spawning aggregations (Claro et al., 2009). In Mexico, there is currently no information about stock structure and gray snappers are caught in a mixed stock fishery (FAO, 2011; SAGARPA, 2012), the health of which is not well known. There are few population genetics studies for the gray snapper and only one study related to genetic connectivity exists (Gold et al., 2009). The authors looked at genetic structure at a scale of 2,400 km in the north of the Gulf of Mexico and the U.S. Atlantic using microsatellites, reporting the existence of genetically different populations and a decrease in their effective size from east to west presumably driven by ocean currents. They recommended that these populations be treated as independent stocks for effective management of the fishery. However, there have been no studies of genetic structure or connectivity of gray snapper in the southern Gulf of Mexico or Mexican Caribbean.

In Mexico, we hypothesize that there may be genetic differences between the populations of gray snapper in the Mexican Caribbean and the southern Gulf of Mexico. Prior studies have shown that there is genetic differentiation between these major basins in other taxa. Blacktip sharks (Carcharhinus limbatus), a low dispersal species, show strong genetic differentiation between the Mesoamerican Barrier Reef System and the southern Gulf of Mexico (Keeney et al., 2005). The bicolor damselfish (Stegastes partitus), a high dispersal reef fish (Hogan et al., 2012), has shown evidence of a weak restriction in gene flow between the Mexican Caribbean and southern Gulf of Mexico (Villegas Sánchez et al., 2014). Similarly, the lionfish (Pterois volitans), the most studied invasive species, has been reported as having significant genetic differentiation between both regions, which suggests a phylogeographic break (Labastida-Estrada et al., 2019). The objective of this study is to determine the diversity and genetic connectivity among seven subpopulations of gray snapper in the southern Gulf of Mexico and Mexican Caribbean Sea.

Materials and Methods

Study area

Two regions were studied: (1) Gulf of Mexico, with the subpopulations Campeche (C), Puerto de Veracruz (PV) and Tuxpan (TX); (2) Mexican Caribbean Sea, with the subpopulations Bahia de Chetumal (BC), Xahuayxol (X), Punta Herrero (PH) and Chiquilá (CH) (Fig. 1). Our seven sampling sites are distributed along approximately 1,950 km of coast between the Caribbean Sea and southern Gulf of Mexico. The Gulf of Mexico region possesses an extensive continental shelf with a diversity of ecosystems like wetlands, the largest area of mangroves in Mexico, coastal dunes and coral reefs with a tropical and subtropical climate (Lara-Lara et al., 2008). The Mexican Caribbean region with a narrow continental shelf is located in the Southeast of Mexico, with a warm subhumid climate and an annual average temperature of 26 °C and a mean annual precipitation of 1,300 mm (Lara-Lara et al., 2008).

Results

Discussion

The main objective of this study was to evaluate stock structure, connectivity, and to estimate genetic diversity among seven subpopulations of gray snapper (L. griseus) in two regions of the Mexican Atlantic: the Gulf of Mexico and the Mexican Caribbean Sea. We found no significant genetic differences between basins or among subpopulations (AMOVA: between regions and among subpopulations). Similarly, estimations of F_ST and R_ST between subpopulation pairs were all not significant. This lack of genetic difference indicates high genetic connectivity among all subpopulations of L. griseus along this ~1,950 km stretch of coastline. This result is also supported by the Mantel Test (p > 0.05), which didn’t show an association between genetic differences and geographic distances. In all cases, the number of migrants per generation (N_m) entering any given subpopulation was greater than four, this indicates that there is unrestricted gene flow and that the populations behave like a panmictic population, as theoretically with the migration of a single organism (N_m = 1) allele fixation is avoided (Slatkin, 1995). However, our estimates of M appear to be greater within basins than between basins, and in the Gulf of Mexico region there appears to be a distinct directionality to migration, with greater M-values between neighboring subpopulation pairs going west compared to east (Table 4; Table S1). This suggests that despite the fact that F_ST based estimates of gene flow show no regionalization, there may be some subtle patterns of reduced connectivity between basins and perhaps a distinct directionality of gene flow in the Gulf of Mexico.

The fact that we found that connectivity may be slightly restricted between the Mexican Caribbean and southern Gulf of Mexico is aligned with findings from previous studies of connectivity between the regions. Blacktip sharks (C. limbatus) are known to show strong genetic differentiation between the Mesoamerican Barrier Reef System (Caribbean) and the southern Gulf of Mexico (Keeney et al., 2005). Additionally, bicolor damselfish (S. partitus) also shows evidence of a weak restriction in gene flow between the Mexican Caribbean and southern Gulf of Mexico (Villegas Sánchez et al., 2014). Our finding here of generally greater migration rates among subpopulations within regions than across regions further supports these studies.

Ocean currents and the biology of L. griseus may play an important role in connectivity among the subpopulations. A particle tracking model of a closely related species Lutjanus analis virtual larvae carried out by Martínez, Carrillo & Marinone (2019) suggests that the marine protected areas of the Mesoamerican Reef network are all highly connected through ocean current. This species has a similar life cycle as L. griseus, therefore the results of this project support the findings by these authors. The Yucatan current (average speed of 1.5 m/s), which later becomes the Loop Current (Athié et al., 2011), can export fish larvae from the Caribbean up to the Gulf of Mexico (Carrillo et al., 2015). Possibly these ocean dynamics favor the dispersion of L. griseus larvae. In adults, migrations are present but these are mainly local, with registered distances between 35 and 122 km (Claro & Lindeman, 2008), and these movements are typically between inshore habitats and shelf habitats rather than among reefs. So it is unlikely that adult movements alone are resulting in high levels of gene flow observed in this study.

In terms of the apparent directionality of gene flow in the southern Gulf of Mexico, studies of oceanographic connectivity in the southern Gulf of Mexico indicate that east-gene flow from Campeche Bank to Veracruz and Tuxpan reefs is low and that connectivity is stronger going west to east (Sanvicente-Añorve et al., 2014). The pattern of migration observed in this study appears to contradict these previous findings (Fig. 2). However, ocean currents in this region are complex with the presence of eddies and a seasonal shifts in direction within the inner shelf with summer months showing flux from east to west and winter months showing flux from west to east (Salas-Monreal et al., 2017). The summer period of greater east to west connectivity also coincides generally with gray snapper spawning season (Domeier, Koenig & Coleman, 1996). The pattern of east-to-west migration observed here is a feasible explanation.

The highest migration rates were between CH–C and C–PV. According to this result, Campeche can be an important point for the connectivity between the two regions. It is important to mention that this Mexican state has the largest extent of mangroves (Lara-Lara et al., 2008), and it has been shown that mangroves are the principal habitats for larvae and juveniles of L. griseus, even though they also use seagrass beds at times (Claro & Lindeman, 2008). For successful sea dispersion, it is essential that larvae find an adequate habitat for recruitment (Cowen, 2006). Thus, if the habitat is fragmented or destroyed, the connectivity can be limited (Jones, Srinivasan & Almany, 2007). We also found high migration rates between subpopulations that were apparently counter to the flow of main currents (e.g., C–CH, CH–BC and PH–X). This could be due to displacing reproductive aggregations that occur between the months of June and August around the new moon on the shelf border with a duration period of 8–10 days (Claro & Lindeman, 2008). Moreover, it has been reported periods of a coastal countercurrent over the shelf (Carrillo et al., 2017) that could promote migrations as it was found in our results.

The effective population sizes estimated from Migrate (θ) were large, however, they must be considered with some caution, especially when interpreting for management and conservation issues. The mutation rate that is, considered for the N_e calculation can create a bias with different values with several orders of magnitude. It is suggested that the values of N_e should only be taken as comparative ones between sampled sites rather than a true point estimate (Beerli, 2016).

Genetic diversity in these populations was high (Ht general average = 0.640; PIC general average = 0.597). According to the PIC, the microsatellites used are highly informative and polymorphic (Botstein et al., 1980). The number of alleles per microsatellite varied between 5 (Lca43) and 15 (Prs260), with a general mean of 6.17 (n = 8); previous studies of L. griseus have reported means of 5.43 (n = 14) (Renshaw et al., 2007) and 7.0 (n = 14) (Gold et al., 2009), similar to the values reported in this study. These subpopulations appear to be similar in diversity to those of the U.S. Gulf of Mexico and southern Florida.

In this study, deviations to the HWE were observed for several loci in some subpopulations with mean values of F_IS varying from 0.066 (X) to 0.188 (TX), which suggests a heterozygote deficit not previously reported for L. griseus. MicroChecker indicated that the most possible cause for the significant values in the F_IS index could be the presence of null alleles. Null alleles are produced when there is a mutation in one of the primer binding sites of the microsatellites; this prevents annealing with the designed primer preventing the amplification of one of the alleles, resulting in a false homozygote (Estoup et al., 1998; Chapuis & Estoup, 2007). Such condition is frequent in microsatellites given their high levels of polymorphism. This has been reported in various species especially in populations with high effective size (Neff & Gross, 2001; Chapuis & Estoup, 2007). Additionally, fish have a high mutation rate in comparison with other classes (reptiles, birds, amphibians and mammals), because they have larger microsatellites and length is an important factor that influences mutation rate (Neff & Gross, 2001). There are ecological explanations for deviations from HWE including inbreeding and genetic drift. The deficiency in this case is unlikely to be explained by inbreeding due to the reproductive habits of L. griseus, given that they form aggregations and in these, gametes are simultaneously released to be fertilized (Claro & Lindeman, 2008). However, heterozygote excesses have been explained by sweepstakes recruitment in marine organisms with small pelagic larvae period and highly variable adult reproductive success (Hedgecock, 1994) which can lead to instantaneous genetic drift.

Conclusions

The high levels of genetic diversity, similar to those observed in other gray snapper populations from the northern Gulf of Mexico, as well as the high levels of gene flow in general, suggest that L. griseus constitutes a single genetic population in the Mexican Caribbean and the southern Gulf of Mexico. The absence of population bottlenecks or disturbances on its connectivity across this large region should be taken into account to define a proper management for the stock. More work is needed to verify the connectivity and the apparent unidirectionality in gene flow in the southern Gulf of Mexico (east-to-west). The next priority for understanding gray snapper populations in the Western Atlantic is to determine the degree of connectivity between Mexican populations and the northern Gulf of Mexico and between Mexico and Cuba.

Supplemental Information