Stoner and Davis: Outplanting queen conch, Strombus gigas 



401 



differed in growth rate (F lw = 10.75, P=0.008). At site 

 C2 there were no differences in growth rate either 

 between hatchery and wild conch or among the stock- 

 ing densities (Table 6). 



Prior to the beginning of the enclosure experiment, 

 pen locations were chosen for similarity in macro- 

 phyte characteristics (Table 7). Seagrass shoot den- 

 sity iF 93 JS =0.55, P=0.938) and macrophyte biomass 

 (F, 



23,48' 



=1.37, P=0.177) did not differ among the 24 

 cages. Although ANOVA showed that detritus dif- 

 fered (F 23 4g =2.00, P=0.022) among the cages, Tukey's 

 multiple comparison test did not detect the differ- 

 ences (P>0.05). At the end of the experiment dry 

 weight of seagrass detritus did not differ among the 

 cages (Table 7)(P 23 ^=0.900, P=0.598), and there were 

 no differences in individual cages between the be- 

 ginning and end of the experiment (P 7 742 =0.090, 

 P=0.764). There is no evidence, therefore, that detri- 

 tus was depleted even at the high density of 2.5 

 conchirr 2 . When comparing shoot density between 

 the beginning and end of the experiment, there was 

 no difference (F, /J9 =2.95, P=0.088), but biomass of 

 living seagrass did differ between the dates 

 (F, ^ 2 =37.01, P<0.001), probably related to blade 

 growth in the spring season. 



Experiment II 



Mortality At the termination (day 85) of enclosure 

 experiment II (29 November 1990-21 February 1991) 



mortality was obviously higher at 

 site C2 than at CI for both hatch- 

 ery and wild conch (Fig. 12), simi- 

 lar to the results of enclosure ex- 

 periment I (Fig. 10). Midway 

 through the experiment there was 

 no significant mortality difference 

 between site CI and C2; however, 

 the difference was significant by 

 the end of the experiment (Table 

 8); mortality was higher at site C2. 

 Similar to enclosure experiment I, 

 there were no differences in mor- 

 tality between hatchery and wild 

 conch (Table 8). 



Growth Growth rates were low 

 in enclosure experiment II (0.01- 

 0.06 mmd" 1 ) (Fig. 13), paralleling 

 the trend observed in free-ranging 

 conch (Fig. 4) and associated with 

 low winter temperatures (Fig. 3). 

 There were significant site x stock 

 type interactions for period I 

 (Fj /2 =5.949, P=0.031) and period 

 II (Fj ;2 =5.004, P=0.045) because 

 of differences in growth rate between hatchery and 

 wild conch at site CI (period I: F t 6 =6.48, P=0.044; 

 period II: F ; 6 =9.747, P=0.021); but not at site C2 

 (period I: F 2 6 =0.008, P=0.932; period II: F 16 =0.2Q7, 

 P=0.665). At site CI wild conch grew approximately 

 twice as fast as hatchery conch. 



Tether experiments 



Experiment I 



Mortality The first tether experiment, conducted 

 from 11 April to 11 July 1990, confirmed that the 

 difference in tag recovery rate between hatchery- 

 reared and wild conch was related to predation (Fig. 

 14). Hatchery conch were killed at a frequency ap- 

 proximately twice that of wild conch for day 45 and 

 day 88 (Table 9, Fig. 14). Site effects were not sig- 

 nificant at either midpoint or end of the experiment 

 (Table 9, Fig. 14). 



Growth Growth rates in both hatchery and wild 

 conch on tethers were higher at site C2 than at site 

 CI by the end of the study period ( 11 July 1990) (Fig. 

 15). This difference also occurred in free-ranging 

 conch (Fig. 5) and enclosure experiment I (Fig. 11). 

 During period I (April and May 1990) there was a 

 significant site x type interaction (Table 10) because 

 wild conch grew faster at site C2 than CI 

 (P /64 =49.28, P<0.001), and hatchery conch grew at 



