Saillant and Gold: Population structure and variance effective size of Lut/anus campechanus in tfie nortfiern Gulf of Mexico 



137 



Sampl 

 in the 

 centra 



suffer reduced capacity to respond to changing 

 or novel environmental pressures (Frankham, 

 1995; Higgins and Lynch, 2001). 



At present, red snapper resources in the 

 northern Gulf are managed under a single- 

 stock hypothesis (GMFMC-^ «). This hypoth- 

 esis is supported by a number of prior genetic 

 studies that employed allozymes (Johnson"), 

 mitochondrial (mt)DNA (Gold et al., 1997; Gar- 

 ber et al., 2004), and microsatellites (Gold et 

 al., 2001b). In each study, genetic homogene- 

 ity was observed across sampling localities, 

 leading to the inference that gene flow was 

 sufficient to maintain statistically identical 

 allele distributions across the sampling area. 

 All of these studies, however, either involved 

 individuals of mixed cohorts or were based on 

 relatively small sample sizes. Alternatively, 

 tag-and-release and ultrasonic tracking (Fable, 

 1980; Szedlmayer and Shipp, 1994; Szedlmay- 

 er, 1997) have indicated that adult red snapper are 

 sedentary and exhibit high site fidelity (but see Patter- 

 son et al., 2001). In addition, Pruett et al. (2005) used 

 nested-clade analysis of red snapper mtDNA sequences 

 and found evidence of different temporal episodes of 

 both range expansion and restricted gene flow due to 

 isolation by distance. They suggested that the spatial 

 distribution of red snapper in the northern Gulf had 

 a complex history that likely reflected glacial advance 

 and retreat, habitat availability and suitability, and 

 that the latter (i.e., physical conditions, and habitat 

 availability and suitability) could partially restrict 

 gene flow among present-day red snapper. 



Our objectives were to more rigorously assess genetic 

 stock structure in the northern Gulf by employing a 

 large sample size of individuals from discrete cohorts. 

 We report allelic variation at 19 nuclear-encoded mic- 

 rosatellites sampled from each of two cohorts at three 

 different localities in the northern Gulf. Genetic homo- 

 geneity among localities was tested and contemporane- 

 ous or variance effective size (A/',,) at each locality was 

 estimated from the temporal variance in allele frequen- 

 cies (Waples, 1989) by using a maximum likelihood 

 method (Wang, 2001). 



33°N 



27°N 



96 W 



84°W 



= GMFMC I Gulf of Mexico Fishery Management Coun- 

 cil). 1989. Amendment number 1 to the Reef Fish Fishery 

 Management Plan. 356 p. Gulf of Mexico Fishery Manage- 

 ment Council, 2203 N. Lois Avenue, Suite 1100, Tampa, FL 

 33609. 



•^ GMFMC (Gulf of Mexico Fishery Management Coun- 

 cil). 1991. Amendment number 3 to the Reef Fish Fishery 

 Management Plan, 17 p. Gulf of Mexico Fishery Manage- 

 ment Council, 2203 N. Lois Avenue, Suite 1100, Tampa, FL 

 33609. 



' Johnson, A. G. 1987. An investigation of the biochemical 

 and morphometric characteristics of red snapper tLutjanus 

 canipecltanus) from the southern United States. Unpubl. 

 manuscr., 26 p. National Marine Fisheries Service, Panama 

 City Laboratory. 3.500 Delwood Beach Road, Panama Citv, 

 FL 32407-7499. 



Figure 1 



ing localities for adult red snapper iLiitJaniis campechanus) 

 northern Gulf of Mexico: northwestern Gulf (Texas), north- 

 1 Gulf (Louisiana), and northeastern Gulf (Alabama). 



Materials and methods 



Adult red snapper were sampled between 1999 and 

 2001 by angling 40-50 km offshore of Port Aransas 

 (Texas), Port Fourchon (Louisiana), and Dauphin Island 

 (Alabama). These localities represent western, cen- 

 tral, and eastern subregions, respectively, within the 

 northern Gulf (Fig. 1) but hereafter for convenience 

 are referred to as Texas, Louisiana, and Alabama. 

 Individual fish were aged by otolith-increment analysis 

 (following Wilson and Nieland, 2001) and individuals 

 belonging to the 1995 and 1997 cohorts were selected 

 for genetic analysis. Sample sizes for the 1995 and 1997 

 cohorts at each locality were 203 and 211 (Texas), 286 

 and 272 (Louisiana), and 376 and 274 (Alabama). Tissue 

 samples (heart and muscle) were removed from each 

 fish and stored as described in Gold et al. (2001b). The 

 genotype of all fish was determined at 19 microsatel- 

 lites by using PCR primers and methods described in 

 Gold et al. (2001b). 



Summary statistics, including number of alleles, allel- 

 ic richness (a measure of number of alleles independent 

 of sample size), and unbiased gene diversity (expected 

 heterozygosity) were computed for each microsatellite in 

 each sample, with F-stat, version 2.9.3 (Goudet, 1995). 

 Homogeneity of allelic richness and gene diversity 

 among samples was tested with Friedman rank tests. 

 Departure of genotypic proportions from Hardy-Wein- 

 berg equilibrium expectations was measured within 

 samples as Weir and Cockerham's (1984) f; probability 

 of significance (Pjjw' was assessed with a Markov-chain 

 method (Guo and Thompson, 1992), as implemented in 

 Genepop (Raymond and Rousset, 1995) and by using 

 5000 dememorizations, 500 batches, and 5000 itera- 

 tions per batch. Genotypic disequilibrium between pairs 

 of microsatellites within samples was tested by exact 

 tests, as implemented in Genepop and by employing the 

 same Markov-chain parameters as above. Sequential 

 Bonferroni correction (Rice, 1989) was applied for all 

 multiple tests performed simultaneously. 



