Statistical Analysis 



lb obtain a single index of stomach fullness, values 

 for the foreguts and midguts were combined by 

 using a weighted average In calculating mean 

 volumes each gut section was assumed to be approx- 

 imately cylindrical. The foregut to midgut volume 

 ratio, determined from five randomly chosen 

 stomachs, was 3.16:1. The following equation was 

 used to calculate the overall gut fullness: 



F X 3.16 + M 

 4.16 



with F and M representing the foregut and midgut 

 values, respectively. 



A multiple regression (General Linear Model- 

 Statistical Analysis Systems) was initially employed 

 to determine which of the variables collected for each 

 fish (i.e, distance upstream, length, weight, sex, and 

 number of repeat spawnings) was most strongly cor- 

 related with stomach fullness, the dependent 

 variable Significant variables identified through 

 multiple regression analysis were further analyzed 

 with chi-square and F-tests. 



Results 



Presence of Food 



Nearly all (91 of 103 or 88%) fish sampled in April 

 1980 contained food (Tkble 1) as did all 15 fish col- 

 lected in April 1981. High percentages of the fish 

 collected on each date in 1980 had food in their 

 stomachs (Tkble 1). About 53% (48 of 91) of the blue- 

 backs in 1980 (Tkble 1) and 73% (11 of 15) of the blue- 

 backs in 1981 had either foregut and/or midgut 



fullnesses of 1/4 or greater. Approximately half of 

 the fish from each date in 1980, with the exception 

 of 13 April, had either foregut and/or midgut full- 

 nesses of 1/4 or greater (Tkble 1). 



Diet 



The diet of the bluebacks collected in 1980 at both 

 sites was composed of zooplankters, benthos, and 

 terrestrial insects (Tkble 2). Chydorid cladocerans 

 were the only zooplankters consumed in large 

 numbers at either location (Figs. 1, 2). Insects, which 

 accounted for 8.1% of the organisms consumed, oc- 

 curred in about half of the fish. Ephemeroptera 

 (Baetis), Coleoptera (Dytiscidae), and Heleidae lar- 

 vae, as well as chironomid larvae and pupae, were 

 the most conspicuous of the identifiable benthic in- 

 sects. Most insects, benthic and terrestial, were 

 unidentifiable. Chironomids occurred more frequent- 

 ly than other insect groups, but they accounted for 

 only 2.7% of the total prey items. Several terrestrial 

 insects were found in stomachs of bluebacks, par- 

 ticularly at Williams' Fishery. Insects, both benthic 

 and terrestrial, increased in importance with time 

 at Williams' Fishery, reaching about 22% during 

 later collections (Fig. 1). Insects represented a 

 smaller proportion of the diet at Rocky Hock Creek 

 (Fig. 2). Fish eggs (probably from alewives or blue- 

 backs), which occurred in the stomachs of several 

 bluebacks in 1980 (Figs. 1, 2), were the most abun- 

 dant food item in that year although their impor- 

 tance decreased with time Varying amounts of sand 

 and detritus occurred in many stomachs. 



In 1981 the diet of bluebacks from Rocky Hock 

 Creek was much less diverse (Tkble 3). Cladocerans, 

 the predominant prey items, comprised 84.1% of the 

 diet. Almost half of the prey items were daphnid 



Table 1.— Incidence of feeding and stomach fullness in male and female blueback herring collected at Williams' Fishery 



(WF) and Rocky Hock Creek (RH) during April 1980. 



712 



