256 



Fishery Bulletin 90(2). 1992 



selective mortality prior to settlement, possibly reflect- 

 ing stochastic variation in the marine environment 

 (e.g., water temperature, salinity), may be responsible 

 for the "chaotic genetic patchiness" that they observed. 

 Alternatively, temporal variation in the source of re- 

 cruits for each locality and/or genetic drift resulting 

 from a finite number of breeders could also generate 

 random genetic patchiness on both temporal and spatial 

 scales (e.g., Waples 1989). None of these aforemen- 

 tioned hypotheses can be excluded with the available 

 data. 



The effective number of breeders (N^) contributing 

 to a cohort of larvae that settle together at a particular 

 location is unknown for S. gigas. Males and females 

 breed in aggregations at characteristic locations over 

 a 6-9 month period, and each female may produce 

 several egg masses of approximately 310,000-750,000 

 eggs each during the breeding season (Robertson 1959, 

 Randall 1964, Weil and Laughlin 1984, Berg and Olson 

 1989). Several females within an aggregation may lay 

 their egg masses simultaneously, and because the rate 

 of embryonic development is temperature-related, 

 hordes of larvae are released synchronously. These 

 larvae can thus be entrained together into the water 

 column and affected simultaneously by marine and 

 oceanic processes. Consequently, hordes of larvae from 

 a finite number of parents could potentially be pre- 

 sented simultaneously to a substrate that would induce 

 settlement and metamorphosis. 



Recently, Bucklin et al. (1989) and Bucklin (1991) 

 obtained evidence that ocean currents and related 

 processes (e.g., upwellings, eddies, offshore jets) can 

 spatially and temporally maintain genetically discrete 

 cohorts of zooplankton in the marine environment. For 

 example, Bucklin et al. (1989) concluded that such pro- 

 cesses "prevented homogenization of the plankton 

 assemblages during transport" and that "plankton 

 populations in complex flow fields may show patchiness 

 in biological, biochemical, and/or genetic character at 

 small time/space scales." Their results suggest that 

 similar processes could affect significantly the distribu- 

 tion of pelagic larvae following their release into the 

 water column. 



The source of S. gigas recruits for the Florida Keys 

 is unknown. The Florida Current, which sweeps east- 

 ward past the Florida Keys and subsequently forms the 

 Gulf Stream, is created by the massive flow of warm 

 water northward from the Caribbean Sea through the 

 Yucatan Channel. This current could entrain large 

 numbers of larvae from numerous locations prior to 

 flowing eastward past the Florida Keys (Mitton et al. 

 1989). Stochastic variations in water currents, surface 

 winds, and meteorological events (e.g., tropical storms) 

 could thus affect significantly the source of S. gigas 

 recruits for any particular locality. During the course 



of our study, we attempted to gain permission to col- 

 lect conch from Cuba and Yucatan, Mexico— two pos- 

 sible sources of recruits for the Florida Keys— but were 

 unable to do so. 



One potential shortcoming of our study was that the 

 temporal effects of recruitment were confounded with 

 other population processes; that is, temporal genetic 

 variation was measured among mixed aggregations of 

 conch sampled in different years at the same locality 

 and not among separate year-classes or cohorts. With 

 the exception of the Coffins Patch population or ag- 

 gregation (see below), all samples consisted of mixed 

 age- and size-classes with new recruits added each year. 

 In addition, some of the temporal genetic variation may 

 have been due to migration of juveniles and adults into 

 and out of the study areas (Hesse 1979, Weil and 

 Laughlin 1984, Stoner et al. 1988, Stoner 1989). Con- 

 sequently, we cannot separate the temporal effects of 

 recruitment from other population processes. How- 

 ever, our goal was not to estimate temporal genetic 

 variation among cohorts or year-classes per se, but 

 rather to provide a measure of within-population (i.e., 

 within-locality) variation by which the significance of 

 genetic variation among localities could be evaluated. 

 Population processes causing temporal genetic varia- 

 tion within localities would similarly affect the genetic 

 variation among localities. Some measure of temporal 

 variation was thus needed before the microevolutionary 

 significance of genetic variation among localities could 

 be ascertained. Alternatively, some form of stratified 

 sampling of year- and/or size-classes would be required 

 to separate recruitment or year-class effects from other 

 potential sources of temporal genetic variation. 



Possible evidence that recruitment, migration, or 

 similar population processes may significantly affect 

 the population structure of S. gigas was the presence 

 of the MDH-2*(138) allele at a frequency of 0.206 in the 

 1987 sample from Ballast Key but the near absence of 

 this allele in the 1988 sample and elsewhere during our 

 study. Mitton et al. (1989) similarly reported, for the 

 Bermuda population, a frequency of 0.30 for a "fast" 

 MDH-2* allele that was also rare elsewhere. However, 

 the Bermuda population is believed to be self-sustaining 

 with little planktonic recruitment from the Gulf Stream 

 or elsewhere (Mitton et al. 1989). Conversely, Ballast 

 Key is situated within the Florida Current and is the 

 most upstream locality from which we collected conch 

 for the present study. Two distinct aggregations of 

 S. gigas may have been sampled at Ballast Key in 1987 

 and 1988, respectively. 



Anomalous results 



Coffins Patch Size distributions suggest that the Cof- 

 fins Patch population was most likely a single year-class 



