Gold et al.: Genetic studies of Scomberomorus cavalla in Florida 



493 



and winter seasons in the "mixing" zone) suggested that 

 2.6-30.9% of recaptured tagged-fish in the Atlantic were 

 returned as Gulf fish and 1.5-13.6% of recaptures tagged in 

 the Gulf were returned as Atlantic fish. These mixing rates, 

 however, were questioned (Jones et al.**) because virtual 

 population analysis (VPA) based estimates of fishing mor- 

 tality for the directed king mackerel fisheries in the Gulf 

 and Atlantic corresponded to annual exploitation rates of 

 0.30 and 0.11, respectively, whereas exploitation rates cal- 

 culated from the 1985-93 (uncorrected) tag returns ranged 

 from 0.027 to 0.033 (Gulf) and from 0.036 to 0.045 (Atlan- 

 tic). The difference between the two estimates of exploita- 

 tion rates implied either that the true exploitation rate was 

 overestimated by VPA or underestimated by uncorrected 

 tag-return data, leading to the conclusion (Jones et al.^) 

 that little confidence should be placed in reported mixing 

 rates based on mark-and-recapture data. 



The goals of this project were to use nuclear-encoded 

 microsatellites to define more rigorously the spatial-tem- 

 poral limits of the two stocks (if separate stocks exist) and 

 to estimate the proportions of both stocks in the mixing 

 zone. The issues of spatial-temporal limits and mixing 

 of the two (presumed) stocks are important in relation 

 to assessing and allocating the king mackerel resource, 

 particularly during the winter season. For example, 

 mark-recapture data (MSAPM indicated that -20% offish 

 tagged in the mixing zone in southeastern Florida moved 

 into the Gulf If this means that only -20% of winter 

 catches from the east coast of Florida are Gulf stock, as 

 opposed to 100% under the current management plan, the 

 allowable biological catch (ABC) for the Gulf stock would 

 decrease significantly (MSAPM. Because the Gulf stock of 

 king mackerel currently is considered overfished (Legault 

 et al.^) reductions in ABC of the Gulf stock could have sig- 

 nificant economic impact. 



The choice to employ microsatellites for the project 

 was straightforward. Briefly, microsatellites are rapidly 

 evolving, short stretches of DNA composed of di-, tri-, and 

 tetranucleotide arrays that are abundant, highly polymor- 

 phic, and inherited in a codominant fashion (Weber, 1990; 

 Wright, 1993; Wright and Bentzen, 1994). Because allele 

 frequencies at microsatellites are generally consistent 

 with equilibrium expectations of diploid, Mendelian loci 

 and because identification of individual microsatellites is 

 by polymerase-chain-reaction (PCR) amplification, both of 

 which remove most of the problems associated with ho- 

 mology of alleles, microsatellites have proven to be useful 

 genetic markers of population structure in numerous taxa, 

 including fishes (Angers and Bernatchez, 1998; Ruzzante 

 et al., 1996; O'Connell et al, 1998; Nielsen et al, 1999). 

 In addition, new alleles at microsatellite loci appear to 

 arise rapidly (Schug et al., 1998), generating high allelic 

 diversity important for statistical power in exact tests and 



^ Jones, C. M., M. E. Chittenden, and J. R. Gold. 1994. Report 

 to the mackerel stock identification working group. Unpubl. 

 document of meeting held 8 Sep 94 to 9 Sep 94 at Panama 

 City Lab., SE Fish. Sci. Cent. 7 p. Southeast Fisheries Science 

 Center, National Marine Fisheries Service, 3500 Delwood Beach 

 Road, Panama City, FL 32408. 



other tests of allclo-distribution homogeneity (Estoup et 

 al., 1998; Ross et al., 1999). 



Materials and methods 



A total of 20 samples of king mackerel were procured be- 

 tween 1996 and 1998 from 11 different offshore sites 

 (Table 1, Fig. 1). The sample from Panama City was 

 obtained from charter boat catches, and the samples 

 from Sarasota and Jacksonville were obtained from tour- 

 naments. The remaining samples were obtained from 

 commercial catches. Tissue samples (heart and muscle) 

 were removed from each fish, frozen in liquid nitrogen, 

 transported to College Station, and stored at -80°C. Sex 

 of individuals was recorded for all samples, except for the 

 March 1997 sample from the Florida Keys. Approximate 

 ages of individuals from all samples except for the July 

 1998 sample from Jacksonville, the March 1997 sample 

 from the Florida Keys, and the April 1997 sample from 

 Sarasota, were determined by otolith-increment analysis 

 by following methods outlined in DeVries and Grimes 

 (1997). 



Initially, we planned to deploy the five microsatellites 

 developed in a prior study (Broughton et al., 2002). Two 

 of these (Sca-8 and Sca-47), however, had proven difficult 

 to amplify consistently in the prior study and therefore 

 were omitted from our study. A third microsatellite, Sca- 

 30. developed by Broughton et al. (2002), also was omit- 

 ted because of difficulties with consistent amplification 

 and because allele distributions at Sca-30 were highly 

 leptokurtic (Broughton et al., 2002). A total of five new 

 microsatellites was then developed from the microsatel- 

 lite-enriched genomic library generated by Broughton et 

 al. (2002). Candidate microsatellites were sequenced from 

 either or both ends by using standard M13 sequencing 

 primers and an Applied Biosystems (Perkin Elmer) 377 

 automated DNA sequencer. The OLIGO software pack- 

 age (National Biosciences, Inc., 1992) was used to identify 

 primers from regions flanking microsatellites. Primers 

 were designed according to preset criteria that included 

 product length, internal stability, proportion GC content, 

 and primer Tm difference. PCR amplifications were per- 

 formed under a variety of experimental conditions to 

 optimize procedures that produced high yields of target 

 sequence and minimized additional fragments ("stutter" 

 bands). Experimental tractability (reproducibility, consis- 

 tency, range in allele size, frequency of "stutter" bands [if 

 present], and microsatellite polymorphism) of PCR-ampli- 

 fied microsatellites were evaluated by screening a panel 

 of king mackerel samples available in the laboratory. PCR 

 primer sequences, the length (in base pairs) of the cloned 

 allele, and the annealing temperature in PCR amplifica- 

 tion for the seven microsatellites used in the project are 

 given in Appendix Table 1. Two of these, Sca-37 and Sca- 

 44, were developed previously by Broughton et al. (2002). 



For assay of individual fish, genomic DNA was isolated 

 from frozen tissues as described in Gold and Richardson 

 (1991). Genotypes at the seven microsatellites were deter- 

 mined by PCR amplification and gel electrophoresis. Prior 



