492 



Fishery Bulletin 97(3), 1999 



level of fishing beyond which the average subsequent 

 recruitment will be reduced. The distribution of stock 

 and recruitment data in this study suggests that 

 maintaining a "spawner threshold" for this stock unit 

 should produce higher average recruitment (Fig. 4). 

 The threshold suggested here, 1.3 billion spawning 

 shrimp is, of course, not an absolute measure, but 

 rather a level that is relevant only to the indices as 

 constructed here, assuming elemental trawl effi- 

 ciency of 0.50. If better information on the average 

 elemental efficiency of shrimp trawl gear can be de- 

 veloped, then an actual spawning threshold can be 

 calculated. With or without an absolute threshold, 

 these data suggest that curtailing harvest in years 

 of recruitment failure has the potential to increase 

 subsequent average recruitment. 



One seemingly surprising finding of this study is 

 that egg production seems to offer little benefit over 

 spawner abundance as an index of shrimp reproduc- 

 tive output. This result is partially an artifact caused 

 by the two indices being based on the same underly- 

 ing estimates of stock area and density, in combina- 

 tion with the use of a static length-fecundity rela- 

 tionship based on pooled data from several samples 

 (Hannah et al., 1995). The egg production index in- 

 corporates only two additional sources of real varia- 

 tion: the variation found in the average size of fe- 

 males and in the percentage of females in the spawn- 

 ing population, and these parameters are not highly 

 variable (Hannah et al., 1995). The finding that egg 

 production estimates offer little benefit over estimates 

 of spawner abdundance is also convenient because the 

 flexible sex change exhibited by ocean shrimp makes it 

 difficult to translate an egg production goal into a quan- 

 tity that can be used to directly manage the fishery, 

 such as a biomass threshold (Charnov et al., 1978; 

 Hannah and Jones, 1991). To do so would require a 

 way to accurately forecast the percentage of both pri- 

 mary females and age-2 males in the fall of the year, a 

 technique not presently available. 



The apparent influence of intense fishing on late 

 egg-bearing females in April may have important 

 implications for management of the ocean shrimp 

 stock and suggests an important interaction between 

 recruitment, parent reproductive output, and varia- 

 tion in the ocean environment. If the timing of the 

 spring transition is indeed the critical determinant 

 of ocean shrimp larval survival, then larvae released 

 early and late in the spring may suffer very differ- 

 ent fates, depending on when the spring transition 

 occurs. When the transition is late, logic suggests 

 that the magnitude of recruitment may depend very 

 heavily on the survival of late release lan'ae, the very 

 ones the fishery is capable of impacting heavily by 

 harvesting egg-bearing shrimp. If this hypothesis is 



correct, it also suggests that environmental varia- 

 tion and the various measures of shrimp reproduc- 

 tive output one might use are confounded. For ex- 

 ample, a large larval release in a year with a late 

 transition may be the functional equivalent of a much 

 smaller larval release in a year with an early transi- 

 tion; both result in low numbers of larvae released 

 into favorable ocean conditions. Similarly, the impact 

 on recruitment of harvesting late egg-bearing females 

 could be large or minimal depending on the timing 

 of the transition. For the 1989 year class, the impact 

 was apparently quite severe, owing to a combination 

 of a large catch of egg-bearing shrimp and a moder- 

 ately high April sea level (Table 1), suggesting a late 

 spring transition. The same amount of harvest may 

 have had minimal impact on recruitment in a year 

 with an early spring transition, when this hypoth- 

 esis suggests that most of the larvae would have al- 

 ready been released into favorable conditions. Upon 

 close examination, the interaction between variation 

 in the timing of the spring transition and the timing 

 and magnitude of larval release makes it unlikely 

 that any simple measure of reproductive output 

 would have direct explanatory power with regard to 

 the level of subsequent reci-uitment. Such a mechanism 

 could also be operating to obscure the stock-reciaiit- 

 ment relationship in other stocks with an early critical 

 stage and environmentally forced recruitment. 



I consider the detection of a stock-recruitment re- 

 lationship for this species to be a preliminary find- 

 ing at this time. If this result endures as additional 

 data become available, the implications for manage- 

 ment of the ocean shrimp stock are still not entirely 

 clear. Regulations aimed at eliminating, or severely 

 reducing, the harvest of egg-bearing female shrimp 

 in April may help increase average recruitment of 

 ocean shrimp, with very little reduction in short-term 

 yield. If the opening date of the ocean shrimp season 

 were delayed 2-3 weeks, the harvest of egg-bearing 

 shrimp in April would be virtually eliminated in ar- 

 eas 84 and 86 (Fig. 6). In more southern areas, like 

 area 88, a full month delay is needed to eliminate 

 harvest of egg-bearing shrimp (Fig. 6). With 6 months 

 of fishing remaining in the open season, it is unlikely 

 that annual exploitation rates would be greatly af- 

 fected by a delay in the season opening date. Although 

 major benefits from such a change in management 

 would not accrue in most years, the evidence for a 

 general stock-recruitment relationship suggests that 

 some small increase in long-term average recruit- 

 ment would result. 



The other novel finding of this study concerns the 

 strong recruitment-stock relationship for ocean 

 shrimp, and its implications for managing highly 

 variable, short-lived stocks. The data show that most 



