c. Immature stage (I) : clearly recognizable germinal 

 follicles are present; male follicles with a thick layer of 

 spermatids around the periphery and a few spermatozoa in the 

 center; female follicles relatively small with peripheral oocytes 

 and occupying rather limited areas of the mantle tissue. 



do Mature stage (M) : Male follicles with or without a thin 

 layer of spermatids but filled with mature spermatozoa; female 

 follicles packed with mature ova; both male and female follicles 

 having displaced most of the Leydig tissue in the mantle; gametes 

 in gonoducts. 



e. Spent (S) : empty or near empty follicles with extensive 

 leucocytic infiltration and phagocytosis. 



To facilitate statistical analyses of the data, the four 

 developmental stages: E, I, M and S were pooled into two 

 categories: the immature stage (I) and the mature stage (M) . 

 The immature stage consisted of individuals classified as in the 

 early development and in the immature stage; the mature stage 

 combined individuals exhibited mature ova and/or contained 

 remnant ova and spent germinal follicles. The frequencies of the 

 two condensed developmental stages: I and M in the four 

 populations are presented in Table 11. The data were subjected 

 to a replicated goodness of fit test (G-test) (Sokal and Rolf, 

 1969) . In the analysis, the ratios of immature and mature 

 individuals at WLISrN, WLISc and 500MW were treated as 

 "replicate" samples which were tested against the ratio of 

 immature and mature (8:73) mussels found at Rlr. 



The analyses show that Gp, Gp and G H are all significant. 

 The three populations of WLISC, 500MW and WLISrN have an excess 

 of immature individuals; the proportions of immature or mature 

 individuals appeared to have been sampled from different 

 populations, in spite of the fact that all three experimental 

 populations were originated from Rlr. The 500MW population has a 

 nonsignificant excess (G=2.958, p>0.05) and the WLISrN and WLISc 

 populations deviate significantly from the expected 8:73 ratio 

 (G=5.968, p<0.025 and G=19.386, p<0.005, respectively). Hence, 

 G T is highly significant. The Gp is also highly significant, 

 because the consistent trend favors immature mussels in the three 

 experimental populations located in the western Sound. The 

 significant heterogeneity G H indicates that the magnitude of 

 favoring immature animals is not uniform in all cases. Based on 

 analyses of the partitioned G values, the ratios of immature to 

 mature mussels in the four populations are not homogeneous; it 

 can also be seen that the heterogeneity of the ratios is 

 contributed by two populations: WLISrN and WLISc, which is 

 verified by the simultaneous test procedure (STP) . These 

 analyses suggest that the proportions of immature mussels at 

 WLISrN (19%) and WLISc (27%) are higher than the expected 

 proportion at Rlr (10%) and are associated with the locations. 



10 



