888 



Fishery Bulletin 97(4), 1999 



Previous studies have reported geographic varia- 

 tion in hfe history traits and population dynamics of 

 Atlantic croaker found north and south of Cape 

 Hatteras (White and Chittenden, 1977; Ross, 1988), 

 suggesting that two groups of croaker with contrast- 

 ing life histories overlap in the Cape Hatteras re- 

 gion, with a mainly offshore group north of Cape 

 Hatteras which displays greater size-at-age, greater 

 longevity, lower annual mortality, and delayed matu- 

 ration in relation to croaker south of Cape Hatteras. 

 A more recent study by Barbieri et al. ( 1994a ) in the 

 Chesapeake Bay concluded that life history varia- 

 tion in M. iindulatits was ephemeral and that the 

 two-stock hypothesis should be re-evaluated. The 

 present mtDNA analysis supports previous sugges- 

 tions by Ross ( 1988) and White and Chittenden ( 1977 ) 

 that life history variation, when present, has an 

 ecophenotypic basis. 



A review of A/, undulatus life history characteris- 

 tics lends support to the panmixia hypothesis for the 

 Atlantic region and provides several mechanisms by 

 which trans-Hatteras gene flow may occur. Mark- 

 recapture studies indicate extensive movement by 

 M. undulatus along the U.S. Atlantic coast and thus 

 potential for gene flow by means of adult migration 

 (Pearson, 1932; Haven, 1959; Bearden, 1964; DeVries^ ). 

 Larval dispersal could augment gene flow. Atlantic 

 croaker larvae are abundant at outer-continental 

 shelf locations (Govoni and Pietrafesa, 1994) and may 

 spend more than 60 days in shelf waters before reach- 

 ing estuaries (Warlen, 1980). Offshore spawning com- 

 bined with an extended pelagic larval stage in shelf 

 waters could provide ample opportunity for larval 

 dispersal among Atlantic localities. Interestingly, 

 spring recruits are occasionally reported in the mid- 

 Atlantic Bight (MAB) estuaries as late as May and 

 June (Haven, 1957; Chao and Musick, 1977) despite 

 reports that spawning ceases in this area by late 

 December (Morse, 1980; Barbieri et al., 1994b). 

 Spring recruits to MAB estuaries may therefore be 

 transported to mid-Atlantic estuaries from spawn- 

 ing sites south of Cape Hatteras. Thorrold et al. 

 (1997) found no significant differences in otolith 

 chemistry of juvenile croaker from the Neuse River, 

 North Carolina, and the Elizabeth River, Virginia, 

 suggesting that immigrants to estuaries north and 

 south of Cape Hatteras may originate from the same 

 spawning area. The reproductive strategy of M. 

 undulatus may also facilitate gene flow. Atlantic 



' DeVries, D.A. 1986. InshoreAtlantic croaker tagging study In 

 J. L. Ross, D. A. DeVries, J. H. Hawkins III. and J. B. Sullivan, 

 Assessment of North Carolina commercial finfisheries, p. 3.31- 

 394. Completion report for Project 2-386-R. North (Carolina 

 Department of Natural Resources and Community Develop- 

 ment. Div. Marine Fisheries. Morehead City. NC. 



croaker are multiple spawners with indeterminate 

 fecundity (Barbieri et al., 1994b). Because spawning 

 occurs during a southerly coastal migration during 

 fall, females could commence spawning within the 

 MAB and spawn multiple times during their south- 

 ward migration, effectively distributing maternal 

 half-siblings (mtDNA clones) over an extensive geo- 

 graphic range. 



MtDNA heterogeneity between Atlantic and Gulf 

 of Mexico localities suggests that these regions sup- 

 port separate populations of M. undulatus. The ob- 

 served genetic break is consistent with a contempo- 

 rary range discontinuity in southern Florida, where 

 M. undulatus seldom occur south of Indian River on 

 the Atlantic coast and are rarely encountered south 

 of Tampa Bay on the Gulf coast (Nelson et al., 1991; 

 Pattillo et al., 1997). Average sequence divergence 

 between Atlantic and Gulf haplotypes (p=0.005'7f) 

 was relatively low, however, suggesting that the phy- 

 logenetic break between Atlantic and Gulf stocks may 

 be less pronounced in M. undulatus than in other 

 marine fishes (Bowen and Avise, 1990). 



This study represents the first attempt to identify 

 population genetic structure in M. undulatus using 

 DNA-level markers. The technique employed in our 

 study proved sufficiently powerful to detect regional 

 (Atlantic vs. Gulf of Mexico) population structure, 

 but did not reveal structured genetic stocks of croaker 

 within the Atlantic. The integrity of Cape Hatteras 

 as a genetic stock boundary could be tested further 

 by using fine-scale markers such as microsatellites, 

 which have revealed genetic differentiation among 

 fish populations that exhibit little mtDNA diver- 

 gence. Alternatively, higher-resolution mutation- 

 screening analysis or direct sequencing of the mito- 

 chondrial D-loop region might provide greater reso- 

 lution of geographic structure than the RFLP tech- 

 nique employed here (Ong et al., 1996; Stabile et al., 

 1996). 



Although mtDNA analysis did not indicate discrete 

 genetic stocks of croaker within the Atlantic region, 

 harvest stocks (see Gauldie, 1988) worthy of man- 

 agement consideration may exist. Because low lev- 

 els of gene flow may produce mtDNA homogeneity 

 between otherwise self-recruiting stocks, mtDNA is 

 incapable of distinguishing between low {17c) and 

 moderate {507f) amounts of mixing (Carvalho and 

 Hauser, 1995 ). Delineation of harvest stocks will thus 

 require more precise estimates of mixing than those 

 obtainable from mtDNA analysis. Mark-recapture 

 studies designed to quantify the level of adult mi- 

 gration across Cape Hatteras, combined with otolith- 

 based microchemical analyses to examine larval dis- 

 persal patterns, could provide valuable information 

 on the level of mixing between the MAB and South 



