580 



Fishery Bulletin 88(3), 1990 



Figure 3 



Number of individuals per sample (2..5-m radius circle) within each 

 conch size-class for each station at Children's Bay Cay, July 1988 

 and February 1989. Size-classes are as follows: A = 40-49 mm, B 

 = 50-59 mm, C = 60-69 mm. D = 70-79 mm. E = 80-89 mm, F 

 = 90-99 mm, G = 100-109 mm, H = 110-119 mm, I = 120-129 

 mm, J = 130-139 mm, K = 140-149 mm, L = 1.50-159 mm. M = 

 160-169 mm, N = 170-179 mm, = 180-189 mm, P = 190-199 

 mm, and Q = 200-209 mm. 



multiple correlation coefficient of 0.664 (;>< 0.001). The 

 stepwise multiple regression model for conch biomass 

 included only shoot density and provided a multiple 

 regression coefficient of 0.513 (7J<0.001). No other 

 variables contributed significantly to the correlation. 

 Results of the regressions and the Newman-Keuls 

 tests for conch density across stations indicate that sta- 

 tions 3-5 provide a suitable habitat for conch larger 

 than 70 mm SL with habitat characteristics ranging 

 as follows: 256-864 mean shoots/m-, 13.9-80.6 mean 

 g dry wt of ThalcL-isialrn", and 0,64-157.12 mean g dry 

 wt of detritus/m-. In addition, the correlation between 

 conch and macrophyte characteristics are linear only 

 up to station 5 (608-864 mean shoots/m-, 62.4-80.6 

 mean g dry wt of Thuldssia/m-, 64.0-157.1 mean g 

 dry wt of detritus/m-) which appears to be the near- 

 optimal habitat for juvenile conch larger than 70 mm. 



Conch length-frequency patterns 



The Kolmogorov-Smirnov two-sample test was used to 

 test for significant differences in the length-frequency 

 distribution of conch at different stations. Only stations 



SHARK ROCK 



1 ^-ru 



rl rh n 



n-r-r-n-Th 



r-n-J h II I I 



ru-n 



DE FGH I J KLMNOPQ 



DEFGHI JKLMNO 



Figure 4 



Number of individuals per sample (2.5-m radius circle) within each 

 conch size-class for each station at Shark Rock, July 1988 and 

 February 1989. For size-classes see Figure 3. 



with at least 10 conch were used in the analysis (CBC 

 stations 3-5 in both July and February; SR stations 3-7 

 in July and stations 5-6 in February). 



In July, CBC stations 3, 4, and 5 all had significant- 

 ly different conch length-frequencies (stations 3 and 4, 

 p<0.0001, stations 3 and b, p = 0.015, stations 4 and 

 5, p<0.0001). Mean lengths varied only from 102 to 

 111 mm, and the statistical differences were a result 

 of subtle differences in the shape of the distributions 

 (Fig. 3). In February at CBC, the length-frequency 

 distribution of conch at station 3 was different signif- 

 icantly from that at station 4 {p = 0.004), and the 

 distribution at station 4 was different significantly from 

 that at station 5 (/)< 0.0001). The length-frequencies at 

 stations 3 and 5 were not different (p = 0.062); the 

 mean values and shapes of the distribution were similar 

 (Fig. 3). 



In July, conch length-frequency patterns at SR sta- 

 tions 3, 4, and 5 were not different (stations 3 and 4, 

 p = 0.220; stations 4 and 5, /) = 0.260), and had means 



