From shelf to shelf: assessing historical and contemporary genetic differentiation and connectivity across the Gulf of Mexico in Gag, Mycteroperca microlepis

Abstract

Describing patterns of connectivity among populations of species with widespread distributions is particularly important in understanding the ecology and evolution of marine species. In this study, we examined patterns of population differentiation, migration, and historical population dynamics using microsatellite and mitochondrial loci to test whether populations of the epinephelid fish, Gag, Mycteroperca microlepis, an important fishery species, are genetically connected across the Gulf of Mexico and if so, whether that connectivity is attributable to either contemporary or historical processes. Populations of Gag on the Campeche Bank and the West Florida Shelf show significant, but low magnitude, differentiation. Time since divergence/expansion estimates associated with historical population dynamics indicate that any population or spatial expansions indicated by population genetics would have likely occurred in the late Pleistocene. Using coalescent-based approaches, we find that the best model for explaining observed spatial patterns of contemporary genetic variation is one of asymmetric gene flow, with movement from Campeche Bank to the West Florida Shelf. Both estimated migration rates and ecological data support the hypothesis that Gag populations throughout the Gulf of Mexico are connected via present day larval dispersal. Demonstrating this greatly expanded scale of connectivity for Gag highlights the influence of "ghost" populations (sensu Beerli) on genetic patterns and presents a critical consideration for both fisheries management and conservation of this and other species with similar genetic patterns.

Fig 1. Map of the Gulf of Mexico showing Campeche Bank (CB) and the two regions of interest on the West Florida Shelf (WFS): the North West Florida Shelf (NWFS) and the South West Florida Shelf (SWFS).

Solid line and arrow indicate general location and direction of the Gulf Loop Current and Gulf Stream Current in the region. Image was taken from U.S. Geological Survey, National Geospatial Program website.

Fig 2. Pairwise measures of genetic differentiation among gag populations in the Gulf of Mexico.

Left section of table shows standardized measures of genetic differentiation (φ’_ST) of microsatellite markers below the diagonal and φ_ST-values of all mtDNA data (sequence and indel repeats) above the diagonal. Right section of table shows φ_ST-values for sequence data only and φ_ST-values for mtDNA indel repeats above the diagonal. “*” denotes population significantly different using exact test; “†” denotes populations significantly different using χ2-test. Number of each respective symbol indicates level of significant (e.g. * is p < 0.5; ** and †† are p < 0.01; *** is p < 0.001).

Fig 3. Unrooted minimum spanning network for mtDNA Control Region sequences.

A represents sequence data only; B represents sequence and indel data. Each circle represents a haplotype where the size of the circle is directly proportional to haplotype frequency and pie slices indicate proportions of haplotypes from each region (Campeche Bank—black; South West Florida Shelf—dark grey; North West Florida Shelf—light gray). Hash marks (█) indicate 1 step difference between sequences. Crosses (χ) indicate a 40-bp indel repeat difference. Diamonds (♦) indicate a 9-bp indel repeat difference.

Fig 4. Mismatch distributions of mtDNA haplotypes from pooled Gulf of Mexico samples.

A-C show results from fitting data to a pure demographic expansion model and D-F show results from fitting data to spatial expansion model. A and D are derived from sequence information only, B and E are derived from indel information only, and C and F are derived from both data-types combined. The demographic expansion model was only rejected for A. In all other case, demographic and spatial expansion models were not rejected. Solid line represents observed data; dotted line represents model prediction; dashed line represent 95% confidence interval for model prediction.

Fig 5. Schematics of three migration networks describing relative Θ and Nm-values.

A—Full migration model; B—Symmetric migration model; C—Scenario 1: Campeche Bank as source population; D—Scenario 2: Connectivity matrix based on ecological data; E—Scenario 3: West Florida Shelf as source population. All circle and arrows sizes are, respectively, proportional to estimates of theta and number of migrants per generation (see S4 Table for values). Black circles represent Campeche Bank population; Grey circles represent South West Florida Shelf; White circles represent North West Florida Shelf. Bayes Factor model comparisons showed C to be very strongly supported (>> 150) by the data over all other migration models.