cladocerans (Ikble 3). Few copepods were consum- 

 ed, although they occurred in about half the fish. 

 Ostracods were important numerically and in occur- 



Table 2.— Diets of 103 blueback herring collected at Williams' 

 Fishery and Rocky Hock Creek during April 1980. 



'<0.1%. 



Table 3.— Diets of 15 blueback herring collected at Rocky Hock 

 Creek on 18 April 1981. 



rence, as were daphnid ephippia. Benthic prey, ter- 

 restrial insects, fish eggs, detritus, and sand were 

 rare 



Number of Repeat Spawnings 



About 85% (87 of 103) of the fish in 1980 and 71% 

 (10 of 14) in 1981 had spawned before (Ikble 4). 

 Although some bluebacks in 1980 had spawned as 

 many as six times, most (72%) had spawned only 

 once or twice before (Tkble 4). Almost equal numbers 

 of males (44) and females (43) were repeat spawners 

 in 1980. 



Table 4.— Number of previous years that 103 

 blueback herring collected at Williams' Fishery 

 and Rocky Hock Creek during April 1980 had 

 spawned. 



Relation Between Feeding Activity and 



Length, Weight, Sex, Number of 



Repeat Spawnings, and Distance Upstream 



Sex was the only independent variable that con- 

 tributed significantly to variance in feeding rate 

 (Tkble 5). Female bluebacks fed more actively than 

 males. The full model explained only 21% of the 

 variance in gut fullness but this was significant {F 

 = 2.87, P>F = 0.005, 102 df). The near significant 

 contributions of length and weight to the reduction 

 of variance in the model is believed to have resulted 

 primarily from sex related differences in mean size 

 (x length females = 305.1 mm, x length males = 

 289.4 mm; x weight females = 258.0 g, x weight 

 males = 212.5 g). F^irther analysis of the feeding 

 activity of male and female bluebacks with a chi- 



Table 5.— Summary of the contribution of each of 

 the independent variables to the multiple regression 

 model for data from 103 blueback herring collected 

 in 1980. Type IV F values and probability levels are 

 shown. P < 0.05 significance level used. 



KO.1%. 



718 



