much too large in comparison with the ulti- 

 mate total catch from the year class. A positive 

 relation appears for the next four year classes, 

 however. 



Admittedly, the total catch of these four 

 year classes is partially estimated since they 

 have not yet passed completely through the 

 fishery. The bulk of the catch has already been 

 made from them, however, so any change in 

 the estimates will not materially affect the 

 relation shown. Considering the additional 

 refinement that can be applied to these data, 

 such as weighting the relative abundance 

 indices for tributaries or sections of the coast 

 on the basis of actual contribution to the 

 fishery (which will be possible as the result of 

 contemplated tagging studies), I believe this 

 work is very encouraging and extremely 

 important to the menhaden studies. The total 

 catch data are not yet available for the 1968 

 year class, but the relative index of abundance 

 for this year class (0.68) is noted in Figure 32. 

 This would indicate a potential total catch of 

 about 1.5 billion fish from this year class. 



Another possible estimate of the strength of 

 the year class is a relation between the catch of 

 0-age fish and the total catch of the year class. 

 This relation permits us to go back before our 

 juvenile abundance surveys were undertaken, 

 to when our aging work started in 1955. Two 

 sets of data are plotted in Figure 32 for each 

 year class: total catch of the year class includ- 

 ing 0-age fish and total catch excluding 0-age 

 fish. These 0-age fish are only partially avail- 

 able to the fishery, and their inclusion could 

 mask a relation that might exist for the older 

 fish. However, these data show that the two 

 largest year classes (1964 and 1966) also had 

 the largest catch of 0-age fish. 



To avoid correlating the 0-age group with 

 itself to some extent, I have compared the 

 catch of 0-age fish with the total catch of all 

 other ages for the year class (Fig. 33). It is 

 apparent that the 1956 and 1958 year classes 

 do not fit the relation which is indicated for 

 the other year classes. If we temporarily ignore 

 these two aberrant year classes, there is a 

 significant correlation for the other 10 year 

 classes (r = 0.889, 8 df ), and a linear 

 regression line for these data is plotted on the 

 graph. 



Returning to the 1956 and 1958 year 



CATCH OF 0-AGE FISH I HUNDREDS OF THOUSANDS) 



Figure 33.— Relation between catch of 0-age fish and 

 total catch of all other ages, Atlantic menhaden, 

 1955-66 year-class. 



classes, there must be some valid reason for 

 excluding them from the analysis, or this 

 relation is of doubtful value. Fortunately, if we 

 look at Appendix Figure 1, which shows the 

 weight data for menhaden caught in Chesa- 

 peake Bay, we see that the 1956 and 1958 year 

 classes produced the smallest fish on record 

 during this period (i.e., as 1-year-olds in 1957 

 and 1959 and as 2-year-olds in 1958 and 

 1960). It may well be that these fish were too 

 small as O's to enter the catch in proportion to 

 their abundance. It appears that there is a 

 definite relation between the catch of 0-age 

 fish and the catch of all other ages, by year 

 class, except that this relation must be modi- 

 fied by the size of the fish. I do not plan to 

 pursue this, but hope this prehminary analysis 

 will stimulate additional study. 



To verify our estimates of year-class strength 

 based on: (1) relative juvenile abundance 

 estimates and (2) catch of 0-age fish in the 

 commercial fishery, we can make an additional 

 comparison between the catch of age-1 fish in 

 the fishery and the total catch of ages 0-3, by 

 year class (Fig. 34). I limited the analysis to 

 fish of ages 0-3 so final figures for each year 

 class could be available within 4 years. Only a 

 small percentage of the catch of a year class is 

 from fish over age 3, so these data basically 

 reflect the year-class strength. As the data 

 become available, total catch of the year class 

 could be used in this analysis. Here again a 

 positive relation exists (r = 0.98, P < 0.01), and 

 the low abundance of recent year classes can be 



18 



