Vaughan and Piager: Decline in abundance of Pagrus pagrus off the soutfieastern United States 



361 



MARMAP 



(Honk and l.inel 



0.9 



0.8 



0.7 



0.6 



0.5  



0.4- 



0.3 



0.2 



0.1 • 



MARMP 

 I Exended chevron) 





Figure 3 



Catch per effort (standarciized to series maximum) from the headboat fishery in North and South Carolina 

 (fish caught per angler day, 1972-97); and from MARMAP sampling by gear Ihook-and-line, 1979-97; and 

 extended chevron trap. 1980-97), for red porgy off the southeastern United States. 



with different assumptions about M. Base analysis of the 

 primary catch matrix estimated weights of 0.08 for hook- 

 and-line CPE and 0.92 for extended chevron-trap CPE; 

 analysis of the alternate catch matrix estimated weights 

 of 0.05 and 0.95. When missing values were used in place 

 of zeroes in the hook-and-line index, estimated weights 

 were 0.50 for hook-and-line CPE and 0.50 for extended 

 chevron trap CPE. demonstrating that treatment of zeroes 

 is an important consideration. The main change in results 

 from using missing values was decreased estimates of F in 

 recent years (Fig. 4A). 



In the retrospective analysis, estimates of F generally 

 converged in about .3 to 4 years; convergence in estimates 

 of recruitment took a year or two longer (Fig. 6). A large 

 positive retrospective pattern was evident in full F in the 



most recent year, and a corresponding negative pattern in 

 estimates of recruitment (Fig. 6). The retrospective pat- 

 tern was similar when using missing values for zeroes. 



Yield per recruit and related benchmarks 



Estimates of equilibrium YPR were somewhat sensitive to 

 the value of M assumed, with larger assumed values of 

 M corresponding to smaller estimates of YPR. The highest 

 YPR was obtained during the middle time period (Table 

 6); however, high theoretical values of yield per recruit can 

 result from fishing mortality rates that are not sustainable. 

 The two biological reference points F,„,^ and Fg j (Bever- 

 ton and Holt. 1957; Sissenwine and Shepherd, 1987) were 

 estimated from the yield-per-recruit analysis. Assuming 



