THE TIMING AND SIGNIFICANCE OF DENSITY-DEPENDENT AND 

 DENSITY-INDEPENDENT MORTALITY OF AMERICAN SHAD, 



ALOSA SAPIDISSIMA 



Thomas F. Savoy and Victor A. Creccoi 



ABSTRACT 



We used stock-recruitment, pre- and postrecruitment mortality data for American shad, Alosa 

 sapidissima, in the Connecticut River to estimate density-dependent and density-independent mortality 

 rates at the prejuvenile (age 1-100 days) and postjuvenile (age 101 days to 5 years) phases. Total post- 

 juvenile mortality rates from 1967 through 1982 were dominated by density-independent mortality, with 

 only 6% (^opost = 0.30) of the mean total (Z = 4.85) being ascribed to density-dependent processes. By 

 contrast, 23% {Z^ = 1.13) of the total prejuvenile mortality was compensatory, of which nearly all 

 took place during the embryonic and early larval periods. Egg and early larval mortality rates from 1979 

 through 1987 were positively correlated to June river flows, and inversely related to mean June 

 temperature, whereas mortality rates of all other life stages showed no such relationships. Daily egg 

 and early larval mortality rates from 1979 through 1987 were usually higher and more variable than 

 mortality rates during later stages. There were significant negative correlations between egg and early 

 larval mortality rates and the strength of 1979-87 year classes in the adult stock, whereas mortality 

 rates of late larvae and juveniles were independent of year-class strength. Density-dependent mortality 

 during the egg and early larval stages comprised over 40% of the total mortality at those stages, resulting 

 in the number of midlarvae and juveniles being positively correlated to adult recruitment. These data 

 suggest that year-class strength of American shad in the Connecticut River is established after the egg 

 and larval stages. 



It is generally accepted that year-class strength of 

 most fishes is established prior to the juvenile stage 

 from density-independent (climatic factors) and 

 density-dependent (competition, predation, cannibal- 

 ism) processes (Gushing 1974; May 1974; Goodyear 

 1980). Although density-dependent mortality is 

 believed to be the mechanism that keeps fish popu- 

 lations stable under low to moderate exploitation 

 (McFadden 1977; Gushing 1980), density-dependent 

 processes such as inter- and intraspecific competi- 

 tion, predation, and cannibalism are difficult to 

 measure. This is particularly so for many highly 

 fecund fishes, whose population sizes undergo wide 

 fluctuations that often cannot be predicted by con- 

 ventional stock-recruitment models (Parrish and 

 MacGall 1978; Bakun 1984). Despite the acknowl- 

 edged importance of density-dependent mortality to 

 stock-recruitment theory (Ricker 1975; Ware 1980), 

 few studies have been able to quantify density- 

 dependent mortality, or determine the life stages 

 at which compensation occurs. 



Density-dependent mortality may be confined to 

 the prejuvenile stage (Gushing 1974, 1980) or the 



'Connecticut Department of Environmental Protection, Marine 

 Fisheries Office, Waterford, CT 06385. 



postjuvenile oceanic stage (Peterman 1978, 1982), 

 or may occur throughout the prerecruitment period 

 (Gulland 1965). The testing of these hypotheses has 

 proceeded slowly because of environmental noise 

 surrounding many stock-recruitment relationships, 

 measurement errors associated with recruitment 

 estimates (Ludwig and Walters 1981), and the lack 

 of long-term mortality data on eggs and larvae. 



One species, for which accurate and long-term 

 stock-recruitment and prerecruitment mortality 

 data exist, is the American shad, Alosa sapidissima, 

 an anadromous clupeid that spawns in many Atlan- 

 tic coast rivers (Walburg and Nichols 1967). Pre- 

 vious studies on American shad in the Gonnecticut 

 River have demonstrated that larval and juvenile 

 mortality rates decline with age (Grecco et al. 1983, 

 1986) and that growth and survival rates among 

 discrete larval cohorts are significantly affected by 

 short-term hydrographic and meteorological events 

 (Grecco and Savoy 1985b, 1987a). Although pre- 

 vious stock-recruitment studies of Gonnecticut River 

 shad (Leggett 1977; Lorda and Grecco 1987) found 

 evidence of density-dependent mortality before the 

 juvenile stage, no studies have attempted to esti- 

 mate density-dependent mortality during the egg 

 and larval stages, or quantify density-dependent and 



Manuscript accepted March 1988. 



FISHERY BULLETIN: VOL. 86, NO. 3, 1988. 



467 



