Table 4.— Analysis of covariance of length by time regression 

 by age and sex for alewives and test for nonzero pooled slope. 

 Damariseotta River, 1979. 



'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. 



applied to each observed age frequency against 

 the age-length key expected frequency, for each 

 day sampled, indicated no significant differences 

 between observed and expected age frequencies 

 in the 1980 data. The magnitude of this decreas- 

 ing trend in length at age may vary from year to 

 year and stock to stock, but it is present and 

 should be taken into account when investigating 

 a migratory stock of alewives. 



Acknowledgments 



I thank David B. Sampson for his help and re- 

 view of this paper, Sherry Collins for her assist- 

 ance in data collecting, and James Rollins for 

 drafting the figures. 



Literature Cited 



Table 5.— Analysis of covariance of length by time regres- 

 sions by age and sex and test for nonzero pooled slope. Damari- 

 seotta River, 1980. 



'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 

 'Pooled slope coefficient 



earliest are not only the largest of the migrating 

 stock but also of the age groups. 



The change in length at age is a source of bias 

 in determining the age composition of the ale- 

 wife harvest if a pooled age-length key were 

 used. Westrheim and Ricker (1978) studied 

 biases connected with application of an age- 

 length key to stocks with different age composi- 

 tions. Using an age-length key derived from 

 pooled length subsamples will introduce bias in 

 computing age composition of an anadromous 

 alewife run. The pooled age-length key assumes 

 homogeneity, but lengths at age change through- 

 out the migration period. Such bias may, how- 

 ever, dwell within the range of acceptable error 

 of 5% of the expected age frequencies. Chi square 



Cooper, R. A. 



1961. Early life history and spawning migration of the 

 alewife (Alosa pseudoharengus). M.S. Thesis, Univ. 

 Rhode Island, Kingston, 58 p. 

 KissiL, G. W. 



1974. Spawning of the anadromous alewife, Alosa pseu- 

 doharengus, in Bride Lake, Connecticut. Trans. Am. 

 Fish. Soc. 103:312-317. 

 Libby, D. A. 



1981. Difference in sex ratios of the anadromous alewife, 

 Alosa pseudoharengus, between the top and bottom of a 

 fishway at Damariseotta Lake, Maine. Fish. Bull., 

 U.S. 79:207-211. 

 Watson, J. E. 



1965. A technique for mounting and storing herring oto- 

 liths. Trans. Am. Fish. Soc. 94:267-286. 

 Westrheim, S. J., and W. E. Ricker. 



1978. Bias in using an age-length key to estimate age- 

 frequency distributions. J. Fish. Res. Board Can. 35: 

 184-189. 



David A. Libby 



Department of Marine Resources 

 Marine Resources Laboratory 

 West Boothbay Harbor, ME 04575 



905 



