Butler et al : Biology and population dynamics of Sebastes levis in the southern California Bight 



275 



uncertainty has little affect on our estimate of the ratio of 

 biomass in 1998 and biomass in 1951 (7.4%) or on the gen- 

 eral conclusion that cowcod biomass was low in 1998. 



Offshore fishing grounds 



Over time, the total number of blocks with rockfish effort 

 in the SCB increased (Fig. 12) as the fishery expanded 

 offshore. Catch rates in blocks nearest to shore have 

 decreased to levels that are low in relation to earlier years 

 (Figs. 1 and 13). Thus fishing grounds in the SCB nearest 

 to shore have been most heavily exploited. Areas of highest 

 cowcod abundance and catch rates are now offshore (Figs. 1 

 and 13). 



Northern areas 



Our analysis focused on the SCB where cowcod abun- 

 dance is highest. However, Butler et al. (1999) examined 

 presence-absence data and CPUE for cowcod in triennial 

 bottom trawl surveys on the continental shelf during 1977- 

 98 (Wilkins'). Cowcod were rare in trawl catches north of 

 the SCB and CPUE was generally zero off Oregon and 

 Washington. However, changes in the spatial distribution 

 of positive tows over time indicated that cowcod became 

 more abundant north of the SCB or colonized northern 

 areas after 1986. 



Rebuilding 



Recent recruitment levels are probably not sufficient to 

 sustain or rebuild the SCB cowcod stock at recent or sub- 

 stantially reduced catches levels. Short-term projections 

 (Fig.ll), assuming recruitment at the estimated average 



for the 1990-98 seasons, indicate that cowcod biomass will 

 continue to decline and that fishing mortality rates will 

 continue to increase over the next ten years at constant 

 catch levels > 5 t/yr (Butler et al., 1999). 



Estimates of rebuilding time for cowcod are much longer 

 than the ten-year default time frame used by managers 

 (DOC, 1998). Jacobson and Cadrin (2002) used cowcod as 

 an example in calculating rebuilding times of 30-50 yr 

 for cowcod with F=0, based on a simple logistic surplus 

 production model, but stressed that their "calculations 

 are examples only and not for use by managers." Refined 

 calculations with better models, additional information, 

 and more realistic assumptions about incidental mortal- 

 ity and recruitment (DeVore''^) give estimated rebuilding 

 times that are generally longer than those of Jacobson and 

 Cadrin (2002). Mean generation time (Restrepoet al., 1998) 

 for cowcod is approximately 37 years. Thus, managers' 

 estimates of the time frame for rebuilding SCB cowcod (one 

 generation time plus expected time to rebuild with F=0) 

 may be greater than 87 yr. 



Recreational fishing effort in the SCB directed at rock- 

 fish, and likely to encounter cowcod, remains relatively 

 high. Logbook records indicate that CPFV vessels alone 

 generate 400-600 thousand angler hours of rockfish effort 

 each year (Fig. 12). MRFSS data indicate that CPFV ves- 

 sels constitute about half of the recreational fishing effort 

 during each year off California. Thus, total recreational 

 fishing effort likely to impact cowcod rockfish might be as 

 high as 800-1200 thousand angler hours/yr 



DeVore, J. 2002. Personal commun. Pacific Fishery Man- 

 agement Council, 7700 Northeast Ambassador Place, Suite 

 200, Portland, OR 97220-1384. 



