566 



Fishery Bulletin 97(3), 1999 



be the major agent of mortality operating on bay 

 anchovy in Chesapeake Bay. The gelatinous zoo- 

 plankters M. leidyi anci C. quinquecirrha, are known 

 to be important predators on eggs and larvae of bay 

 anchovy (Feigenbaum and Kelly, 1984; Monteleone 

 and Duguay, 1988; Cowan and Houde, 1993; Purcell 

 et al., 1994). In Chesapeake Bay, the peak periods of 

 bay anchovy spawning and gelatinous zooplankton 

 abundance overlap during June and July ( Cowan and 

 Houde, 1993; Purcell et al., 1994), facilitating the 

 predator-prey interaction. We found gelatinous 

 predator biovolumes to be significantly higher in 

 June, when only M. leidyi was present, than in July 

 when both species occurred. Chrysao/'a quinque- 

 cirrha, a potentially more powerful predator on an- 

 chovy eggs and larvae than M. leidyi (Purcell et al., 

 1994), occurred only in July, but at low mean 

 biovolumes that were uniform in the three Bay re- 

 gions. It is worth noting that C. quinquecirrha is also 

 a predator on M. leidyi (Purcell and Cowan, 1995), 

 resulting in a predator-prey interaction that poten- 

 tially has a sparing effect on anchovy eggs and lar- 

 vae (Cowan and Houde, 1992). 



The large biovolumes of gelatinous predators prob- 

 ably contributed to the greater mortality rates in 

 June compared with July. However, the six mortal- 

 ity coefficients from the regional estimates were not 

 significantly correlated with gelatinous predator 

 biovolumes (Table 4), despite the strong, negative, 

 linear relationship between anchovy larval abun- 

 dance and gelatinous predator biovolume (r^=0.75; 

 Fig. 8A) for the combined June and July regional 

 data. This negative correlation may have been a con- 

 sequence of predation, but it also could have been 

 generated by lower egg production of anchovy in ar- 

 eas where jellyfish were abundant, as Dorsey et al. 

 (1996) hypothesized in site-specific studies of bay 

 anchovy egg and yolksac larval mortality. Although 

 we cannot conclude unequivocally that gelatinous 



predators accounted for high mortality rates, it is 

 likely that they were significant consumers of an- 

 chovy larvae. 



Abundance data illustrated in survival curves 

 (Figs. 4 and 5) showed several modes, which sug- 

 gested that cohort-specific mortality might be vari- 

 able on shorter time scales than we studied and might 

 be changing as a function of ontogeny, age, or size. 

 Other factors also could have biased our mortality 

 estimates or contributed to regional variability in 

 rates, for example, pulses of spawning that produce 

 variable initial abundances of daily cohorts or immi- 

 gration and emigration of larvae into and out of a 

 region. Because mortality is size-specific, the pres- 

 ence of larger larvae in July could have led to a lower 

 estimate of mortality rate during that period. But, 

 when only larvae of equivalent lengths (i.e. <13 mm 

 SL) were analyzed, estimated mortality rates re- 

 mained nearly twice as high in June (M=0.41) as in 

 July (M=0.22). If some larvae were being advected 

 up the bay, as Dovel (1971) had hypothesized, this 

 process could have contributed to biased estimates 

 of higher mortality rates in the lower bay. 



Despite a high cumulative mortality rate, the lower 

 bay in July had the highest production of larvae sur- 

 viving to 18 days after hatching for the June-July 

 1993 period (Table 3). This result is due to high 

 spawning activity and initial concentrations of lar- 

 vae in the lower bay ( Rilling and Houde, manuscript 

 in review), a high growth rate of larvae, and, impor- 

 tantly, the relatively large volume of water in the 

 lower bay, which supported a large contingent of 

 anchovy larvae. 



Survival and recruitment potential of anchovy co- 

 horts were responsive to variability in both mortal- 

 ity rates and growth rates that they experienced. The 

 ratio M/G, an index of stage-specific mortality, is an 

 important indicator of comparative production and 

 survival potential during early life (Houde, 1996, 



