MANZER: DISTRIBUTION AND FOOD OF STICKLEBACK 



Table 3. -Equivalent units of important dietaries. 



Organism 



Bulk units 



Organism 



Bulk units 



was reasonably uniform with time, seasonal ad- 

 justment of equivalent units appeared unneces- 

 sary. Items which averaged less than one per 

 stomach or less than 1.0% of the bulk were record- 

 ed as trace (T) quantities. 



Stomachs of large stickleback frequently con- 

 tained several hundred organisms. In such cases, 

 contents were identified and enumerated from a 

 weighed portion of the total bolus and the result- 

 ing counts were then prorated to the total weight 

 to estimate the numbers of organisms consumed. 

 The remaining portion of the bolus was examined 

 for food organisms not represented in the sample. 

 Correlation analysis indicated a very significant 

 positive relationship between actual and estimat- 

 ed counts of major food items (r = -i- 0.964, P<0.01, 

 n = 15). 



Major features of the stickleback diet were 

 adequately described from examinations of 10 

 stomachs per sample. In a few cases, smaller 

 numbers were examined to eliminate gaps in time 

 or place. On the basis of two separate tests of 

 association between stomach contents of 10 and 25 

 fish samples from the same catch, ranked by 

 numbers, Spearman's rank correlation test (Siegel 

 1956) gave r, values of -1-0.943 and 1.000. The 

 extent of lake coverage in the 2 yr, especially 1970, 

 differed between surveys. The dietary agreement 

 among stickleback taken at different locations 

 within surveys was examined using Kendall's 

 coefficient of concordance test (Siegel 1956). For 

 each survey, the most common food items at each 

 location were ranked according to mean number in 

 the sample, excluding material rendered uniden- 

 tifiable through digestion. Corrections were made 

 for items tied in rank and W, the index of diver- 

 gence of observed from perfect agreement, and 

 related chi-square values were calculated. For 

 eight of the nine surveys tested (four in 1970 and 

 five in 1971) the agreement observed in rankings 

 of dietaries among locations was higher than it 

 would be by chance {P = 0.05) (Table 4). Therefore, 

 it seemed reasonable to combine the data for all 

 locations by survey to facilitate detection of 



possible seasonal changes in diet. From plankton 

 studies conducted in Great Central Lake in 1970, 

 LeBrasseur and Kennedy (1972) stated that "the 

 epilimnion is well mixed, thus assuring a nearly 

 uniform dispersal of planktonic organisms along 

 the lake." 



Diet in relation to sexual maturity was deter- 

 mined from combined samples of stickleback 

 caught during the first three surveys (mid-May to 

 early June) in 1971. Mature and immature females 

 were separated on the basis of size, 60 mm being 

 used as the dividing length. Of 54 females 60 mm 

 or larger examined, 4 were immature and 50 were 

 mature. Of the latter group, 28 were ripe. Blue 

 coloration of the iris and red coloration of the 

 pelvic region were used to separate mature from 

 immature males (Craig-Bennett 1931; Greenbank 

 and Nelson 1959). Because female sticklebacks are 

 larger than males of equivalent age (Greenbank 

 and Nelson 1959; van Mullem and van der Vlugt 

 1964) males larger than 60 mm were considered to 

 be sexually mature. From testes inspection, ripe 

 males were few in number and accordingly no 

 attempt was made to treat functional and non- 

 functional males separately. The relative scarcity 

 of ripe males is believed due to their behavior of 

 attending spawning females or nests. 



Diel Feeding Rhythm 



Diel feeding periodicity and food composition 

 studies were based on paired catches made at 

 station 1 on 1-2 October 1970, and 21-22 July 1971 

 at two sites (A and B), approximately 100 m apart. 

 In the October series, fishing started at 1300 h 1 

 October and during the next 24-h period was 

 conducted at 1600, 1900, 2200, 0630, and 1000 h. 



Table 4.— Summary of results of Kendall coefficient of concor- 

 dance ( WO tests (Siegel 1956) for agreement in diet of threespine 

 stickleback at different sampling locations. 



655 



