shore from the sampling area and their growing 

 ability to avoid capture. As a consequence, the 

 sequence of these low average monthly lengths 

 cannot be construed to indicate the rate of growth 

 for these young shad. It was held desirable, 

 therefore, to estimate the position of certain points 

 and of the curve in the preparation of figure 8. 

 No precise quantitative justification can be given 

 for these estimates, but they were based on broad 

 sampling experience and extensive observations 

 on catches in inshore waters. The estimates 

 are held surely to be superior to judgment based 

 on the sadly biased sample records. The curve 

 and points of figure 8 beyond age-group O are 

 based strictly on the records of table 11. The 

 records of monthly growth of table 12 were read 

 from the curve of figure 8. 



Gizzard shad of age-group O presumably 

 grew most rapidly in July, August, and September 

 and had peak growth in August. The record 

 for I- and II-group shad indicates that growth 

 in length of the I-group fish progresses rapidly 

 earlier in the year than it does for the Il-group. 

 Greatest gain occurred a month earlier — in July. 

 Growth was slight for both groups during May 

 and for the I-group after September. 



During the second year of life the male shad 

 grew in length 57 percent as much as during the 

 first year, and in the third year, 28 percent as 

 much. For the female shad the second- and third- 

 year growths were 66 percent and 34 percent, 

 respectively, of the first-year growth. 



Seasonal changes in weight resemble those in 

 length (compare figs. 8 and 9) except that weight 



decreases over winter from the peak attained dur- 

 ing the previous fall. The basis for this change 

 was explained earlier in the section on seasonal 

 changes in weight. 



Attempts to follow seasonal growth of gizzard 

 shad by use of tagged individuals were unsuccess- 

 ful. No fish were recaptured from some 600 tagged 

 during the winter of 1952. 



CALCULATED GROWTH 



Calculated Lengths of Age Groups 



The dependability of calculations of length at 

 time of annulus formation is indicated by com- 

 parison of the calculated lengths and the lengths 

 of shad which were captured between the end of 

 the growing season and the time of annulus forma- 

 tion (first and last columns of table 13). Because 

 the annuli are the only landmarks on the scale 

 which can be related to a somewhat definite time, 

 and because the shad has practically the same 

 length from the beginning of the calendar year to 

 the time of the annulus formation, the length of a 

 shad captured in this period should agree closely 

 with the length calculated from an older fish to the 

 corresponding annulus. 



Agreement was close among adequately repre- 

 sented age groups except for the I-group shad. 

 In the compilation of the empirical data for this 

 group, 105 shad had to be eliminated because sex 

 could not be determined with certainty; their 

 average standard length was 129 mm. Had they 

 been included, the disagreement between empirical 

 and calculated lengths would have been less, but 



Table 12. — Progress of season's growth in length of gizzard shad by millimeters and percentage 



[Lengths taken from the curves of figure 8] 



1 These fish are assumed to be about a week old. Shad hatched in the laboratory were 3.5 mm. in standard length and became 5.2 mm. long in 4 days. 

 GIZZARD SHAD IN WESTERN LAKE ERIE 405 



