9-8 



(1974-1977) and c) mortality. The analysis was carried out with respect 

 to total number of oysters, live oysters and dead oysters. Total number 

 of oysters reflected the number known to be alive, the number known to 

 be dead and the number of oysters with missing tags but known to be 

 alive. Missing oysters vrere considered dead except when untagged 

 oysters were not noted, (as in 1975) . In this latter case, all missing 

 oysters were excluded from the analysis as were oysters whose status was 

 questionable. Tests of this hypothesis used a per experiment alpha 

 level of 0.05. 



Hypothesis (2) net change in length (growth) varies by 

 station or by year, was tested by means of a covariance analysis (Guen- 

 ther, 1964) . The covariate or "concomitant" variable in this analysis 

 was initial length and the dependent variable v/as final length. The 

 independent classification variables were stations and years. Data were 

 treated as a 2-f actor, completely randomized design with one concomitant 

 variable. Scheffe's method for multiple comparison (Scheffe, 1969) was 

 employed when so indicated by the results of the ANOVA (Analysis of Vari- 

 ance) . Additionally, because of problems of heterogeneity of variances 

 (see statistical appendix) , three independent covariance analyses were 

 used to assess station variation within years. Factors were a) stations 

 and b) initial length (covariate) . No transformation was applied to the 

 data set because none were consistently successful in normalizing the 

 variance for the entire data set. Data included all oysters for which 

 initial and final lengths were recorded. 



Hypothesis (3) length changes at the two experimental 

 stations occur in the same time periods for a given year, was tested 

 with a mixed model randomized block ANOVA for each station within each 

 year (Kirk, 1968; Gill, 1977). Tukey's procedure for multiple compari- 

 sons was used to assess mean length differences by months . When vari- 

 ances were heterogeneous, modified Tukey's procedures were used. Com- 

 parison of growth patterns were assessed based on differences in results 

 for the two stations. Data included all oysters that were measured in 

 all months. 



