370 



Fishery Bulletin 100(2) 



3000 



2000 



° 1000 



1000 2000 3000 



Total Spawning Stock Biomass (nit) 



4000 



CL 4000 -\ 



3000 



2000 



1000 







500 1000 1500 2000 2500 3000 3500 4000 4500 



Total spawning stocl< biomass (t) 



Figure 9 



Total spawning-stock biomass and rt'cruitniont. to age 1 for 

 red porgy off the southeastern United States from calibrated 

 VPA based on lA) primary catch matrix, and (B) alternate 

 catch matrix. Solid lines represent recruitment predicted from 

 Beverton-Holt models; dashed lines run from the origin with 

 slope of median of recruits divided by the spawning-stock bio- 

 mass for 1972-94 spawning years. 



able and allows calculation of total population fecundity at 

 Z and M, we recommend exploring that approach. Popu- 

 lation simulations for a protogynous grouper (Huntsman 

 and Schaaf 19941 suggest greater vulnerability to fishing 

 than for a comparable gonochoristic stock. It appears that 

 the questions of what SPR measure is most appropriate 

 and how it should be computed cannot be answered until 

 the overall reproductive biology of the species is far better 

 understood. 



As usual for fish populations, the spawner-recruit rela- 

 tionship is only roughly defined by the data (Fig. 9), and 

 the years in which the model does not fit raise interesting 

 questions. In the fit based on the primary catch matrix 

 (Fig. 9A), the two large residuals at the start might be 

 taken as suggestive of a dome-shaped recruitment curve, 

 they simply might reflect random error, or they might re- 

 sult from the additional computations needed to extend 

 the catch matrix back to 1972. The analysis based on the 

 alternate catch matrix (Fig. 9B) shows a more pronounced 

 pattern, in which data from the 1970s demonstrate higher 

 recruitment for a given level of spawning-stock size than 

 later data. This pattern might be taken as evidence of a 

 regime shift (change in underlying ecological conditions, 

 from natural or anthropogenic causes), or it might be an 



artifact of the smaller data sets of age-length data used in 

 this approach. Other explanations are equally possible for 

 residual patterns in Figure 9, A and B. These obsei-ved da- 

 ta suggest "depensation"( reverse compensation;decreased, 

 rather than increased, recruitment per spawner at lower 

 population sizes) at very low spawning biomasses, but this 

 suggestion is uncertain because of high retrospective error 

 in estimating recruitment in the most recent years. 



Growth overfishing was not estimated to be severe. 

 Mean full F in the early period was estimated well below 

 both F,,,,,. and F,, ,. Mean full F in the middle period was 

 estimated as about \'a2'7( of F,, j, and mean full F for 

 the recent period was estimated as about VdVc of F^ j, 

 both below F^,,,^ (Table 7). This stock appears to be one 

 in which recruitment overfishing can be induced before 

 growth overfishing, and for that reason reference points 

 such as Fj,^ j^ and even F,, , would be insufficiently conser- 

 vative for its management. 



The estimate of B^yg^ computed from the recruitment 

 model and selectivity vector can be an appropriate target 

 biomass for management and the corresponding MSST can 

 be a threshold to define overfished status (Restrepo et al., 

 1998). Estimates of spawning-stock biomass were well above 

 MSST, 1972-78; approached MSST. 1982-86; and were well 



