Rilling and Houde: Variability in growth and mortality of Anchoa mitchilll 



559 



rates could be related to biological or environmental 

 factors. The four independent variables considered 

 for inclusion were gelatinous predator biovolume 

 (mL/m-^), zooplankton density (number/L), tempera- 

 ture, and salinity. Mean regional estimates of inde- 

 pendent variables were entered into the regression 

 model. Prior to multiple regression analysis, simple 

 correlation analyses were run to determine which 

 independent variables might be colinear and unsuit- 

 able for inclusion in the multiple regression model 

 (SAS Institute, 1990). Pairs of independent variables 

 with a correlation coefficient >0.70 were considered 

 highly correlated and thus excluded from the mul- 

 tiple regression analysis. 



significantly between June and July 1993 (^-test, 

 P<0.0001). In each month, there was a well estab- 

 lished pycnocline. Baywide mean temperature in- 

 creased from 25.3°C in June to 26.6°C in July Mean 

 salinity baywide increased from 10.1 psu in June to 

 15.9 psu in July (Table 1) and increased in both 

 months between the upper and lower regions of the 

 Bay (ANOVA, P<0.001). Mean dissolved oxygen lev- 

 els decreased from 7.8 mg/L in June to 6.3 mg/L in 

 July. In June, 3-m DO levels ranged from 4.1 to 12.6 

 mg/L ; in July they were lower, 4.6 to 8.4 mg/L. Hy- 

 poxic and near-anoxic conditions (<2.0 mg O2/L) were 

 most prevalent in the deep channel of the mid bay 

 region, especially in July. 



Zooplankton analyses 



In the laboratory, zooplankton organisms were iden- 

 tified by using a dissecting microscope. Zooplankton 

 that were potential prey of larval anchovy were enu- 

 merated, i.e. copepods, barnacle nauplii, gastropod 

 veligers, bivalve veligers, cladocerans, rotifers, 

 tintinnids, polychaete larvae, and chaetognaths. Cope- 

 pods were categorized as adults, copepodites, and nau- 

 plii. Densities of zooplankters were calculated as 



D=NIV. 



where D = density of organisms (number per liter); 

 N = number of organisms in a 10-L Niskin- 



bottle sample: and 

 V - sample volume ( 10 liters). 



Densities were weighted according to the 

 depth range represented by each sample to ob- 

 tain a weighted mean density in the entire wa- 

 ter column: 



* 



1=1 



D 



m 



; = 1 



where m = depth ranges represented by each 

 Niskin-bottle sample: and 

 k = number of Niskin-bottle samples 

 taken on a CTD cast (3 or 4). 



Results 



Hydrography 



Mean temperatures, salinities, and oxygen lev- 

 els at 3-m depth in Chesapeake Bay differed 



Growth 



Baywide, mean growth rate of anchovy larvae in- 

 creased from 0.59 mm/d in June to 0.72 mm/d in July 

 (ANCOVA, P<0.001) (Fig. 2). When the analysis in- 

 cluded only larvae <13 mm SL, an even higher bay- 

 wide growth rate was estimated in July (0.78 mm/d) 

 than in June (0.59 mm/d) (ANCOVA, P<0. 001). Esti- 

 mated lengths at 15 days after hatching were 11.65 

 mm SL in June and 13.72 mm SL in July. 



Growth rates did not differ significantly among 

 regions in June or July (ANCOVA, P>0.05) (Fig. 3). 

 The regional estimates of growth-in-length rates 

 ranged from 0.53 mm/d in the lower bay during June 

 to 0.78 mm/d in the upper bay in July. Although not 

 significantly different, the highest regional growth 

 rates for each month were in the upper bay. The 



30 

 25 



20 



J, 15 

 I .0 



10 



15 20 



Age (days) 



25 



30 



35 



Figure 2 



Baywide linear growth models for bay anchovy larvae in Chesa- 

 peake Bay, June and July 1993. L = standard length (mm), d = 

 age in days, estimated from otolith-increment analysis.  = June, 

 = July 



