Age distribution from the 1979 and 1980 samples 



Age: 3 1+5 6 7 8 9 Total 



1979 



Males 5 134 79 11 4 1 - 234 



Females 1 115 116 22 3 1 1 259 



1980 



Males — 116 221 22 2 — — 361 



Females 2 93 211 32 6 — — 344 



Age by time regressions for the 1980 data were 

 computed for each sex and an analysis of covari- 

 ance(Table3) revealed no significant differences 

 between the slopes. The pooled regression like 

 the pooled length by time regression, showed a 

 significant slope (—0.02) at the 5% level. These 

 two nonzero slopes (length by time and age by 

 time) are evidence of changes in mean size and 

 mean age throughout this alewife migration. 

 The slope of the length trend (—0.36) is greater 

 and significantly different from the age trend 

 slope (—0.02) at the 5% level. The daily age com- 

 position of alewives moving to their spawning 

 ground is a result of the fish schooling by 

 size. 



Lengths were separated into respective age 

 categories for both years and regressions were 

 computed of the age-length relation over time for 

 each sex (Fig. 2). This was not done for ages three 

 and seven through nine because of the small sam- 

 ple size. Ages four, five, and six were predomi- 

 nant, constituting over 95% of all the fish in 

 the samples. All regressions produced negative 

 slopes, although regressions A and B in 1979 and 

 C and D in 1980 (Fig. 2) proved their slopes to be 

 nonsignificant from a zero slope. 



An analysis of covariance applied to all regres- 

 sions in each year showed no significant differ- 

 ences between slopes (Tables 4, 5). Apparently 



Table 3.— Analysis of covariance of age by time regressions 

 for male and female aiewives and test for nonzero pooled slope. 

 Damariscotta River, 1980. 



1 F test, ratio of mean square of difference between slopes to mean 

 square of difference within slopes. 

 Calculated probability, not significant at the 5% level. 

 3 Pooled slope coefficient. 



325 



320- 



3I5- 



3IO 



— 305 



x 

 I- 

 o 



i 2 ! 300 



I979 

 -2^2!H£I3 N-22 



-® 6^ 



O 



295 



290 



285 



280 



_i 1_ 



15 20 25 30 35 



x 



H 



o 



10 15 20 



DAYS 



25 



30 



35 



Figure 2.— Regressions of length at age with time from the 

 1979 and 1980 Damariscotta River alewife fishery. 



the nonsignificant slopes still contributed to the 

 homogeneity of the combined slopes producing 

 significant 1979 and 1980 pooled regression 

 slopes of —0.24 and —0.21, respectively. These 

 results show, along with the previously observed 

 decrease in length with time and age with time of 

 this alewife migration, that the length of fish at 

 age also decreases with time. The fish that arrive 



904 



