SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD 



of the regression differed significantly from zero, 

 indicating the presence of significant density- 

 dependent mortaHty during the prejuvenile stage. 

 When prejuvenile mortality rates were related to 

 both egg production (Eggs,) and mean June river 

 flow, (JFLOW,), the multiple regression model ac- 

 counted for 80% of the variability in prejuvenile 

 mortality (Table 3), and the slope estimates for egg 

 production (6) and June flow (c) were positive and 

 highly significant. Note that the standard error (SE) 

 about the slope estimate (6) was reduced by 60% 

 when June flow effects were considered. These 

 results suggest that prejuvenile mortality rates are 

 affected by a combination of density-dependent (egg 

 production) and density-independent (June river 

 flow) factors. 



The mean density-dependent mortality rate dur- 

 ing the prejuvenile phase (Z^pj.^ = 1.18) from the 

 multiple regression model was four times greater 

 than the mean Z^post value (0.30) for postjuveniles 

 (Table 3), suggesting that 80% (1.18/1.18 + 0.30) 

 of the total compensatory reserve for American shad 

 occurs before the juvenile stage. Whereas compen- 

 satory density-dependent mortality may play a sig- 

 nificant role in regulating egg and larval abundance 

 during years of high egg production, it is clear that 



most of the variability (82%) in egg and larval abun- 

 dance is ascribed to density-independent factors. 



Egg, Larval, and Juvenile 

 Mortality, 1979-87 



Analysis 



To examine how larval and juvenile mortality 

 rates varied with year-class strength, survivorship 

 curves for 19,000-180,000 American shad larvae 

 and 800-3,500 juveniles were developed annually 

 from 1979 to 1987 following the techniques of Lough 

 (1976) and Hewitt et al. (1985). Mortality rates could 

 not be estimated directly for prolarvae because 

 American shad yolk-sac larvae (7-9 mm TL) remain 

 in deep water (Marcy 1976) and were only partially 

 susceptible to the plankton seine. 



All larvae and juveniles collected annually were 

 separated into four length intervals: 1) 10-13 mm 

 TL, reflecting first- feeding larvae with undeveloped 

 pelvic fins (Wiggins et al. 1984); 2) 14-19 mm TL, 

 associated with the onset of pelvic fin development 

 (Lippson and Moran 1974); 3) 20-28 mm TL, repre- 

 senting larvae approaching metamorphosis, char- 

 acterized by invagination of the gut (Maxfield 1953); 



9.0 



ao 



W 7.0 



6.0 



6.0 i 



T- 

 20 



— r- 

 30 



— r- 

 40 



I 

 60 



— I— 

 60 



— r- 

 70 



I 

 80 



— I— 

 90 



100 



no 



120 



130 



Total Eggs (bMons) 



Figure 4.— Relationship between American shad prejuvenile mortality (ZEJ^) and the number of eggs produced 



(Eggs,) from 1967 through 1987. 



473 



