696 



Fishery Bulletin 89[4), 1991 



weight at 80 °C to constant weight and incineration at 

 500°C for 4 hours. Organic content was quantified as 

 the percent difference between dry weight and ash-free 

 dry weight. Another subsample of approximately 50 g 

 was analyzed for granulometric properties. The sam- 

 ple was washed to remove salts and to extract the silt- 

 clay fraction (<62^m) which was analyzed with stan- 

 dard pipette procedures (Galehouse 1971). The sand 

 fraction (>62^m) was analyzed with standard sieve 

 fractionation procedures (Folk 1966). Product-moment 

 statistics were generated for mean grain-size and 

 sortedness (McBride 1971). 



Analysis-of -variance techniques were applied accord- 

 ing to Sokal and Rohlf (1969), with log u > transforma- 

 tions where Bartlett's test indicated heterogeneity of 

 variances (macrophyte data). Conch growth data pro- 

 ved to be heterogeneous in variance in most cases, as 

 indicated by Bartlett's test and plots of means versus 

 variances. Heterogeneity was improved little by log- 

 transformation. Additionally, necessary replacement 

 of conch during the transplant experiment produced 

 a violation in independence in the growth data. For 

 these reasons, Kruskal-Wallis analysis of variance was 

 followed by nonparametric Mann-Whitney U-tests for 

 multiple comparison of growth rates. 



Results 



Changes in habitat characteristics 



Juvenile conch were transplanted to four different 

 habitat types based upon macrophyte biomass: zero, 

 low-, moderate-, and high-biomass habitats (Table 1). 

 Sediment grain-size varied from 1.15 to 2.56 <)> at the 

 eight sites; however, there were no significant dif- 

 ferences in mean values (ANOVA, F 2.26, P>0.05), 

 and all are classified as fine to medium sand. Similar- 

 ly, there were no differences in sediment sorting coef- 

 ficients (F 2.09, P>0.05), with all sites except the bare 

 sand (C3) in the poorly sorted range. Significant dif- 

 ferences occurred in sediment organic content (F 15.97, 

 P< 0.001) with highest values at the high-biomass site 

 (L2) and lowest at the sand site (Table 1). There were 

 no significant differences in sediment organics among 

 the sites with low or moderate biomass. 



At the beginning of the experiment there were no 

 significant differences in biomass of either seagrass or 

 macrodetritus between the two cages at any site (Stu- 

 dent's (-test, P>0.05). Analysis of variance indicated 

 that macrodetritus varied significantly (F 76.48, P< 

 0.001) among the eight sites, but there were no signifi- 

 cant differences in mean values for the three low- 

 biomass stations (Nl, L3, Wl) or for the three moder- 

 ate biomass sites (Cl, C2, LI) (Neuman-Keuls test, 

 P>0.05). Macrodetritus was more abundant at the 



