log f= 7 2227 + 0I76FL 

 IHighom and Nicholson) 



V 200 



250 275 300 

 FORK LENGTH (mm) 



Figure l— Regression of fecundity on fork length for Atlantic 

 menhaden showing confidence limits on the mean at 95'? level 

 ForlogF = 7.2227 + 0.0176^,^ = 38, SE regression = 0.3069; 

 SE regression coefficient = 0.001 1, r^ = 0.726; for logf" = 8.6463 

 + 0.012QFZ,, Af = 101. SE regression = 0.3330, SE regression 

 coefficient = 0.0007, r^ =0.871 



for larger fish. Higham and Nicholson { 1964), e.g., 

 had four fish with between 400,000 and 500,000 

 ova, four with between 500,000 and 600,000, and 

 one with over 600,000, whereas from nearly three 

 times as many fish I had only two with over 

 500,000 and six with between 400,000 and 

 500,000. I believe, however, that differences in 

 counting techniques caused the differences in ova 

 estimates. I used proportional sampling, whereas 

 they did not. I separated the eggs from each other 

 and from the connective tissue, dried and weighed 

 the eggs, and then counted those in a sample. They 

 counted the eggs in a sample from the wet ovary. 

 Also, a certain amount of subjectivity is involved 

 in distinguishing between maturing and non- 

 maturing ova. 



Reproductive Potential and 

 Net Reproductive Rate 



Since the sex ratio of Atlantic menhaden is 

 about equal (Nicholson and Higham 1964), I was 

 able to calculate the annual numbers of females of 

 each age in the population, 1955-68, by dividing 

 half of the number offish caught at each age by the 

 exploitation rate for all ages (Schaaf and 

 Huntsman 1972). When I collected my material in 

 1970, recording the maturing stage offish in catch 



samples had been discontinued, but Higham and 

 Nicholson (1964) estimated that about 10'7f of 

 age-1 fish, 909c of age 2, and lOf/i of age-3 or older 

 fish examined during the North Carolina fall 

 fishery in October-December from 1955 to 1959 

 had maturing ovaries. From these figures I calcu- 

 lated the number of females of each age that would 

 spawn each year and multiplied it by the mean 

 number of ova spawned by fish of each age to 

 estimate the number of eggs spawned each year 

 (Table 2). 



The net reproductive rate, R„,oi' a population is 

 defined as the sum of the products of the age- 

 specific survival rate 4, and the age-specific natal- 

 ity rate w,, of females (Odum 1971). A value of 1.0 

 for each generation would indicate that the popu- 

 lation is stable and that there is a balance between 

 births and deaths. In fish populations it is nearly 

 impossible to obtain accurate counts or estimates 

 of the number of offspring produced by each age- 

 group. It is possible, however, to estimate the 

 mean number of eggs spawned for fish of each age. 

 If this variable is used for m, in the formula given 

 by Odum and ifi/,"!, is called, /?o*. then the recip- 

 rocal of /?„* should be a rough estimate of the 

 survival rate of female eggs, providing the popula- 

 tion is approximately stable. Although the Atlan- 

 tic menhaden population declined after about 

 1960. 1 think in view of the imprecise estimates of 

 other parameters, that it can be assumed stable for 

 the purpose of estimating egg mortality. 



Ro* values were calculated for the 1954-63 year 

 classes. I assumed a 0.65 survival rate up to age 1, 

 which I divided into the estimated number offish 

 that were age 1 (Schaaf and Huntsman 1972) for 

 an estimate of the number of fish at the postlarval 

 stage. Since the sex ratio is equal, this number 

 divided into the estimated number offish surviv- 

 ing to each age iSchaaf and Huntsman 1972) 



Table 2. — Estimated number of eggs (multiply by 10"! 

 spawned by Atlantic menhaden, by year and age. 



310 



