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



263 



Department of Fish and Game plays a key role because 

 cowcod are caught primarily off southern California. 

 Cowcod received relatively little attention from managers 

 until Butler et al.'s (1999) stock assessment indicated that 

 the SCB stock was "overfished" and that "overfishing" was 

 occurring. The Magnuson-Stevens Fishery Conservation 

 and Management Act and National Standard Guidelines 

 (DOC, 1998) require each FMP to specify a minimum stock 



size threshold (B 

 rate threshold iF^ 



Threshold 



), and a maximum fishing mortality 

 ,). According to Council guidelines 



(PFMC, 1999),fi„,^^^,^_,,,y for cowcod is 25'X of virgin biomass. 

 ^ Threshold-^ 40", '^he fishing mortality rate that reduces 

 spawning biomass per recruit to 40% of the unfished level; 

 Clark, 1991) when stock biomass is at or above 40% virgin 

 biomass. F. 



Threshold 



is reduced at lower biomass levels. 

 According to the National Standard Guidelines, a stock 

 is overfished when stock size falls below B 



Threshold 



and 



overfishing occurs when fishing mortality rates exceed 

 ^Threshold ^°^ ^ period of One year or more. The goal for most 

 rebuilding plans is to achieve the target biomass level 

 (usually Bj^gY' the stock biomass for maximum sustained 

 yield) in ten years or less. However, even with zero fishing 

 mortality, ten years may not be sufficient to rebuild some 

 overfished stocks, and this is the case for cowcod. In such 

 situations, the National Standard Guidelines allow for a 

 rebuilding time no longer than one mean generation time 

 plus the expected time to recovery in the absence of fishing 

 mortality. 



In this paper, we summarize existing and new informa- 

 tion (Butler et al., 1999) about cowcod; develop an extended 

 time series of catch and abundance index data; estimate 

 biomass, fishing mortality, and recruitment since 1951 



(with crude but plausible estimates for 1916-50); describe 

 current status of the stock and effects due to fishing and 

 environmental changes; and discuss problems and oppor- 

 tunities in rebuilding the stock to higher abundance levels. 

 In addition, we show how standardized abundance indices 

 can be derived from rare event and presence-absence data. 

 Finally, we demonstrate techniques for tuning stock assess- 

 ment models to presence-absence indices with binomial 

 distributions, low expected values, and zero values. 



Materials and methods 



We estimated annual commercial landings for cowcod 

 during 1951-97 from two different types of information. 

 Commercial landings estimates during 1980-97 were 

 from the PacFIN (PSMFC-^) database based on exvessel 

 sales receipts (total landings by market category) and port 

 samples (used to estimate proportions of each species by 

 market category). During the period of 1980-97, cowcod 

 comprised 0.5% of total commercial rockfish landings in 

 California. The time series of annual ratios for cowcod and 

 total rockfish landings was variable and showed no clear 

 trend over time. 



Direct estimates of cowcod landings were not available 

 for years prior to 1980 because no port sampling was 

 conducted to partition the catch for rockfish market cat- 

 egories into species-specific components. Consequently, we 

 used the ratio estimate (0.00479, CV=26% ) to reconstruct 

 historic annual cowcod landings from 1916 to 1981 based 

 on total reported rockfish landings in California in CMAS- 

 TER records (California Commercial Fisheries Data Base, 



