AHRENHOLZ ET AL.: ATLANTIC MENHADEN POPULATION AND FISHERY 



LU 

 O 



z 



LLI 





EFFORT 



Figure 16. — Differences between observed Atlantic menhaden population 

 fishing mortality rates (Fp,ipl and their mean plotted against nominal effort 

 by fishing season. 



puter program MAREA (Epperly et al. 1986) as- 

 suming exponential growth of the biomass in the 

 population (Epperly and Nelson 1984). This pro- 

 gram, modified for Atlantic menhaden from 

 MGEAR (Lenarz et al. 1974), uses the Ricker- 

 type yield-per-recruit model (Ricker 1975) and 

 permits estimation of yield in each of the Atlantic 

 menhaden fishing areas, with area-specific 

 growth and fishing mortality rates. Each com- 

 puter run of the model generates a matrix of yield 

 per recruit at varied age of entry and multiples of 

 F for each of the fishing areas (5) and one for the 

 entire fishery. Yield for the entire fishery can be 

 obtained either by summing yield from each area 

 or by calculations based on input from the entire 

 fishery estimated as a unit. A summation of the 

 five matricies should be similar (but not equal) to 

 the independently calculated entire fishery ma- 

 trix. Input for the model includes estimates of 

 area specific proportional fishing mortalities, es- 

 timates of weight at age by area and for the entire 

 fishery, and an estimate of M (0.45). 



Estimates of area-specific quarterly F at age 

 were obtained by apportioning F at age for the 

 entire fishery (obtained from the quarterly VPA, 

 1965-78 year classes) with the ratios of the num- 

 ber at age caught in a given area to the total 

 number at age landed in the entire fishery. CThe 

 resulting proportional F's are not equivalent to 

 true area specific F's, but are the correct values 

 for the yield-per-recruit model used). 



Area-specific length at age estimates for the 

 beginning of each quarter for each year class were 



rearranged to correspond to fishing season, and 

 converted to weight at age using season-specific 

 weight-length equations (from Table 4). Parallel 

 conversions were done on lengths at age for the 

 entire fishery to estimate weights at age for the 

 fishery as a whole. 



Annual trends in historic yield per recruit were 

 examined with the fishing mortality, age and 

 size, and geographic pattern extant during each 

 year from 1970 to 1978. Results from these com- 

 putations indicate a severe decline in actual yield 

 per recruit for the entire fishery and Chesapeake 

 Bay area from 1971 to 1978 (Fig. 17). 



Estimates of potential changes in yield per re- 

 cruit under regimes of varied age at entry and 

 (multiple) changes in fishing mortality rate were 

 obtained by averaging parameters reflecting con- 

 ditions during 3-yr intervals, i.e., 1970-72, 1973- 

 75, and 1976-78 (Table 9). Attainment of the max- 

 imum potential yield from Atlantic menhaden in 

 the purse seine fishery would have required a 

 very high rate of fishing at a substantially de- 

 layed age at entry of about 3 years of age 

 (Figs. 18, 19). More practically, yield could have 

 been increased by reducing F's. For example, 

 with a F multiplier of 0.6 and the current age of 

 entry, the gain would have been 6.99f for the con- 

 ditions of 1976-78 (Fig. 19, Table 9). With an in- 

 crease in age of entry to 1.0 (eliminate the harvest 

 of age-0 fish) the gain would have been 10.2%. 

 The patterns of potential gain for conditions 

 under the 1970-72 and 1973-75 time periods are 

 similar, but of lesser magnitude (Fig. 18, Table 9). 



589 



