726 



Fishery Bulletin 103(4) 



The two readings performed to estimate the measure- 

 ment error were not statistically different as indicated 

 by the parameters of the regression line fitted to the 

 first measurement versus second measurement data 

 (SL 1 = 0.992 SL 2 + 0.21, /- 2 = 0.998). The slope was not 

 statistically different from one (regression slope=0.992, 

 SE = 0.005; H : slope = l; ?-test of regression slope, 

 P=0.106), and the intercept was not statistically differ- 

 ent from zero (regression intercept=0.21; SE = 0.12; H : 



•10 10 20 30 40 50 60 70 80 90 



Time of preservation (days) 



Figure 1 



Change in standard length of Atlantic menhaden (Brevoortia 

 tyrannies) (n=100) during 90 days of preservation in 95% 

 alcohol. Mean values and standard error of length measure- 

 ments obtained from repeated measurements of 100 fish. 



y-intercept=0; r-test of regression intercept, P=0.418). 

 Measurement precision, the absolute values of the dif- 

 ference between the two series of length measurements 

 of the same specimens averaged 0.12 mm (SD = 0.09), 

 which corresponded to an average of 0.47% of length 

 (SD = 0.35). Thus, the error associated with measure- 

 ment is an order of magnitude less than the change in 

 length due to shrinkage within the first three days 

 of preservation. Changes in length during following 

 87 days were below measurement error. 



Discussion 



This research on late-larval and early-juvenile Atlan- 

 tic menhaden shrinkage is the first for this spe- 

 cies. Maillet and Checkley (1991) used a shrinkage 

 correction formula (cited as unpubl. data) in their 

 study on larval menhaden growth but did not provide 

 additional information (e.g., range offish sizes) to 

 accompany their formula. Their correction formula 

 differs from ours, and the discrepancy may be related 

 to differences in experimental procedure and differ- 

 ent developmental stages. In the present study live 

 fish were used, but in Maillet and Checkley 's study 

 (1991) it was not indicated whether larvae were alive 

 or dead prior to preservation. Further Maillet and 

 Checkley (1991) examined larval menhaden (17-24.5 

 mm SL), whereas we examined late-larval to early- 

 juvenile menhaden (19.1-31.4 mm SL). 



The shrinkage of larval and early-juvenile Atlan- 

 tic menhaden after the first three days of preserva- 

 tion was significant, but small in magnitude. Be- 

 yond 20 days of preservation significant additional 

 shrinkage did not occur. In fact, the length changes 

 after day 3 were below the estimated measurement 



19 21 23 25 27 29 



Live standard length (mm) 



31 



33 



19 21 23 25 27 29 31 



Preserved standard length (mm) 



33 



Figure 2 



Length changes of Atlantic menhaden (Brevoortia tyrannus) during preservation for 90 days in 95% alcohol (;i = 100). (A) 

 The relationship between live standard length (LSL) and relative (%) shrinkage magnitude; (B) the relationship between 

 live and preserved standard lengths described with linear regression. The solid line indicates the 1:1 ratio. The arrow 

 points to the correction curve obtained from Maillet and Checkley (1991): SL / „,,. = 0.978(SL ic 1 - s " 



