FISHERY BULLETIN; VOL, 72. NO. 4 



Table 4. — Analysis of variance of multiple regression of percentage mortality 

 on temperature and time for embryos and larvae of Mercenaria mercenaria. 

 *** — significant at the 0.001 level; d.f. — degrees of freedom; MS — mean squares. 



indicated by the values of the standard partial 

 regression coefficients (Table 5). However, for 

 cleavage stages, in the final predictive equation 

 T^ became relatively less important (Snedecor, 

 1956; Steel and Torrie, 1960). 



The 88 residuals were tested for skewness 

 (g^) and kurtosis (g^) (Sokal and Rohlf, 1969). 

 Both statistics were normally distributed for 

 cleavage stages (g^^ = -1.02; ^2 "" -1-78). They 

 were as follows for trochophore larvae: g^ = 

 2.02*; g^ = 0.90 and for straight-hinge larvae: 

 g^ = -0.12;^2 = 5.92** (* P = 0.05; **P = 0.01). 



We estimated temperature levels for 10%, 50%, 

 and 90% mortality for each period of time 

 exposure by plotting percentage mortality from 

 Tables 1 to 3 against log temperature on prob- 

 ability paper. These values (Figure 4) allow us to 

 estimate the possible effects of temperature eleva- 

 tion over time on the survival of the different 

 stages. 



DISCUSSION 



Early cleavage stages of molluscs appear to 

 have a narrower range of tolerable temperatures 

 than older stages (Pelseneer, 1901; Loosanoff, Mil- 

 ler, and Smith, 1951; Loosanoff and Davis, 1963; 



Goodwin, 1970; Kennedy et al., 1974). Results for 

 Mercenaria mercenaria indicate that increased 

 temperature tolerance occurred as early as the 

 trochophore stage. This is in agreement with our 

 results for Mulinia lateralis (Kennedy et al., 

 1974). 



Cleavage stages of Mercenaria mercenaria were 

 generally more temperature sensitive than those 

 of Mulinia lateralis (Kennedy et al., 1974). 

 Trochophore larvae of both species were generally 

 similar in their thermal tolerances. Straight- 

 hinge larvae of the hard clam were more tempera- 

 ture tolerant. 



The sensitive cleavage stages of the hard clam 

 are of primary importance in relation to the 

 effects of entrainment and exposure to high 

 temperature in cooling systems. If the cleavage 

 stages are killed, obviously it does not matter 

 that the next stages would be more temperature 

 tolerant. The hard clam spawns during the 

 summer throughout its geographical range (e.g. 

 Loosanoff, 1937; Landers, 1954; Carriker, 1961; 

 Porter, 1967; Chanley and Andrews, 1971). 

 However, little precise information exists as to 

 the temperature range for spawning. Carriker 

 (1961) found that hard clams spawned between 

 22° and 30°C in Little Egg Harbor, N.J., with 

 maximum frequency over the range of 24° to 26°C. 

 This is in general agreement with the sparse 



Table 5. — Statistics of multiple regression of percentage mortality on temperature and time. M = minutes; 

 T = °C; b = regression coefficient; Sf, = standard error of fe; lOOR^ = coefficient of determination (increases 

 as each new variable is added); Srs - standard error of the response surface (decreases as each new variable 

 is added); b' = standard partial regression coefficient (absolute values). 



Developmental 

 stage 



Constant 



Variable 



lOOfl' 



Cleavage 

 stages 



Trochophore 

 larvae 



Straight-hinge 

 larvae 



-3.6 



152.6 



8.7 



1164 



