Dorsey et al.: Cohort abundances and mortality of Anchoa mitchilli 



263 



020 



15 



N 



g 10 



o 



E 



a> 



CO 



LU 



005 



20 40 60 80 



Distance from mouth of bay (nautical miles) 

 Figure 3 



Instantaneous mortality rates (Z/h) of bay anchovy 

 eggs collected at seven sites in Chesapeake Bay, 

 July 1991. Locations of sites are illustrated in Fig- 

 ure 1. Vertical bars represent ±1 SE. 



compared with sites farther north (Table 2; Fig. 1). 

 Observed variability in mortality rates was approxi- 

 mately equal for eggs and yolk-sac larvae. Coeffi- 

 cients of variation (ratio of standard deviation to 

 mean) were 0.73 and 0.79 for eggs and yolk-sac lar- 

 vae, respectively. 



The regression coefficients, which were the esti- 

 mates of mortality rates, did not differ significantly 

 from zero at the a = 0.05 level in five of the twelve 

 log p egg abundance-on-time regressions and eight of 

 the twelve yolk-sac larva regressions. The yolk-sac 

 larval regressions were based upon relatively few 

 data points, which contributed to their lower preci- 

 sion (Table 2). The nonsignificant (P>0.05) mortal- 

 ity coefficients corresponded to low, near-zero regres- 

 sion coefficients. Although not significantly different 

 from zero, these coefficients were presumed to esti- 

 mate mortality rates. 



Correlations 



There were several simple correlations among organ- 

 ism abundances, biovolumes, and mortality rates 

 (Table 3). Most notable were 1) the positive correla- 

 tion between yolk-sac larval mortality rate and C. 

 quinquecirrha biovolume (P<0.01); 2) the positive 

 correlations between egg abundance, yolk-sac larval 

 abundance, and zooplankton biovolume (P<0.05); and 



3) the negative correlations between both egg and 

 yolk-sac larval abundances and M. leidyi biovolume 

 (P<0.05). 



Discussion 



Daily cohorts of bay anchovy were demonstrated to 

 suffer high and variable mortality during egg and 

 yolk-sac larval stages. Our estimates of mortality 

 rates indicate both day-to-day and regional variabil- 

 ity in Chesapeake Bay. At site 5, which was sampled 

 on five consecutive days, the range of egg mortality 

 rates exceeded the range observed over all remain- 

 ing sites. The high variability at a single site sug- 

 gests that environmental conditions were changing 

 daily. Variability in initial egg abundances indicated 

 that numbers of adult spawners in the area also were 

 changing on a daily basis. 



Our egg-stage mortality rates ranged from 2% to 

 98%, and averaged 73%. Recent experiments in free- 

 drifting mesocosms (Houde et al., 1994) conducted 

 at the same sites and dates of surveys reported here 

 have also provided estimates of mortality of bay an- 

 chovy eggs and yolk-sac larvae. The mesocosms, 

 which included gelatinous predators, provided mean 

 estimates of mortality that were essentially identi- 

 cal to those reported here. In Great South Bay, New 

 York, Castro and Cowen (1991) reported that sea- 

 sonal mortality rates of combined bay anchovy eggs 

 and yolk-sac larvae ranged from 70% to 98%/d. If eggs 

 in Great South Bay hatched in 20 h, the mortality 

 rates on eggs alone may have ranged from 63.0% to 

 96.5%/d. In Biscayne Bay, Florida, Leak and Houde 

 (1987) compared seasonal abundances of eggs and 

 yolk-sac larvae of bay anchovy and estimated that 



Table 3 



Matrix of correlation coefficients for bay anchovy, Anchoa 

 mitchilli, where Z is hourly egg mortality rate, ZL is hourly 

 yolk-sac larvae mortality rate, N is initial abundance of 

 eggs, NL is initial abundance of yolk-sac larvae, CTV is 

 biovolume in mL/m 2 of M. leidyi. CHV is biovolume in mL/m 2 

 of C. quinquecirrha, ZP is zooplankton biovolume in mL/m 2 . 

 Asterisks indicate significant correlation (P <0.05). 



7.1 



N 



Nl 



CTV CHV 



-0.49 



