Stoner and Ray Strombus gigas in a marine fishery reserve 



555 



shape, including shell length, are induced by habi- 

 tat characteristics of juvenile queen conch ( Martin- 

 Mora etal., 1995 (and are conserved in adults (Stoner 

 and Schwarte, 1994). For example, juvenile conch that 

 develop in shallow seagrass beds of the bank cease to 

 grow in shell length at a smaller size in comparison 

 with conch developing on the shelf in Exuma Sound. 



To test for differences in the density of adult queen 

 conch among depth zones and between sites, 2-way 

 analysis of variance ( ANOVA) was performed on log- 

 transformed data (log K1 (/2 + l): there were numerous 

 zeros in the data) in accordance with the guidelines 

 of Day and Quinn ( 1989). Prior to analysis, variances 

 were tested for heterogeneity with Cochran's test 

 (P>0.05). The 2-way ANOVA was followed by Tukey's 

 HSD multiple comparison test. 



Because the comparison between the MFR near 

 WW and the fished area near LSI constitutes an 

 important aspect of this study, basic results from LSI 

 (modified from Stoner and Schwarte, 1994) are re- 

 ported in some figures and tables. 



Juvenile surveys 



To estimate abundance of juvenile queen conch at 

 the two sites, we calculated the areal extent of ju- 

 venile aggregations at both WW and LSI. Aggrega- 

 tions of juvenile conch have abrupt boundaries and 

 high densities are found adjacent to habitats with- 

 out conch (Stoner and Ray, 1993). To estimate the 



size of aggregations in the study areas, divers were 

 towed systematically over these areas to determine 

 the boundaries within which density was >0.1 conch/ 

 m 2 . Boundaries were marked with buoys, and buoy 

 positions were determined with GPS. The potential 

 error in positions measured with GPS is variable and 

 can be as large as 150 m. However, when the points 

 were plotted on a small-scale chart, they character- 

 ized accurately the shapes defined by the buoys in 

 the field, and any error was uniform in direction and 

 magnitude. Absolute position was not critical for the 

 calculation of aggregation sizes. Surface areas of the 

 aggregations were determined with the plots and a 

 digitizing board. Surveys of juvenile queen conch 

 have been conducted each summer since 1988 at LSI 

 and were conducted in 1991 at WW. Juvenile conch 

 were infrequently encountered on the island shelf; 

 consequently, densities there were determined in 

 conjunction with the adult survey. 



At LSI, the juvenile surveys were concentrated in 

 the flow field that begins just north of the island and 

 runs south of Norman's Pond Cay onto the Great 

 Bahama Bank and in the flow field that runs between 

 Children's Bay Cay and Rat Cay (5 km) (Fig. 1C). 

 Although most conch were associated with these tidal 

 channels, the entire bank between Rat Cay and the 

 north end of LSI was searched. At WW, the survey area 

 comprised the flow field south of the island which ex- 

 tended onto the bank toward the south end of Hall's 

 Pond Cay (6 km) and the surrounding bank areas (Fig. 



