Vaughan and Prager Decline in abundance of Pagiw, pagni'i off the southeastern United States 



359 



headboat catche.s are ,shown for comparison, but were not 

 used in calibrating VPAs,) 



In the primary catch matrix, modal age (underlined in 

 Table 4) varied between 2 and 5; age 4 has been most com- 

 mon in recent years. Modal age in the alternate catch ma- 

 tri.x (Table 4) was similar, but showed less annual varia- 

 tion, probably because of interpolation across years of the 

 age-length keys. A few strong year classes have moved 

 through the population, most notably those that were age 

 2 in 1987-90. but no strong year classes have followed. 

 The modal age has increased since 1992, the year in which 

 a 12-inch minimum size limit was imposed (amendment 4 

 ISAFMC-'^]). 



Trends in mortality and recruitment 



In our calibrated VPAs, ages 4 through 7 were always con- 

 sidered fully recruited, and estimates of F were typically 



averaged over those ages, weighted by population num- 

 bers at age. Those weighted averages are referred to here 

 as full F. 



In analysis of the primary catch matrix, estimates of F 

 for all ages were lowest in the early years (Table 5). Fish- 

 ing mortality on ages 1-3 increased between the early and 

 middle periods and then declined, but full F (ages 4-1- ) in- 

 creased and remained high. This pattern probably reflects 

 the imposition of the 12-inch TL minimum size limit in 

 1992 with amendment 4 (SAFMC"'). Estimates of F from 

 the alternate catch matrix in the recent period were gen- 

 erally higher for ages 3-1- than estimates from the prima- 

 ry catch matrix. Slightly lower estimates of F in the re- 

 cent time period were obtained from the primary catch 

 matrix when missing values, rather than small numbers, 

 replaced zeroes in computing the hook-and-line index. 



Estimates of full F from both catch matrices started low, 

 rose abruptly from the late 1970s to 1982, decreased some- 



