FISHERY BULLETIN: VOL. 70, NO. 3 



10 



Figure 6. — Within-season changes in catchability coef- 

 ficient (g). Line connects seasonal mean values. Circled 

 points were not used to compute means. 



I I 1 1 ' — 



o o Monthly Volues 



Igl- • ■• Seosonol Averoges 



Average 



I 



(XIO'l 



n \ T" 



MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY 

 55 5€ 57 58 59 60 61 62 63 64 65 66 67 68 



MONTH AND YEAR 



Figure 7. — Monthly catchability coefficients (q). Dashed 

 line shows 1955-1968 mean. 



Table 8. — Estimated percentage of numbers of shrimp 

 functioning as spawners at age I. 



models for predicting recruitment from popu- 

 lation biomass were tried. 



Model I : Ri+2 = aSi e-^^i ; 



Model II: Ri+2 = cSi e-^Pi+i 



where 



Ri + 2 = 



Si = 



Pi+i = 



number of age I shrimp on May 1, 

 season i + 2. 



average biomass of functioning fe- 

 males during September of sea- 

 son i. 



average total biomass (ages I, II, and 

 III) during September of season 



a, b, c, and d are constants. 



Model I assumes the number of eggs produced 

 is proportional to spawning biomass and that 

 survival from egg to recruitment is influenced 

 by this same spawning biomass. Model II as- 

 sumes the number of eggs is proportional to 

 spawning biomass and that survival from egg to 

 recruit is a function of average biomass com- 

 peting for the population space. September of 

 season i + 1 was selected for Model II because 

 this seemed likely to be proportional to the av- 

 erage biomass encountered by age shrimp, and 

 data were available for all Septembers. May 1 

 was selected for recruitment since most seasons 

 opened on this date. 



Both models of recruitment were fitted by 

 using transformations and a linear model (Paul- 

 ik and Gales, 1965) . The transformed equations 

 are: 



Model I : \oge{Ri+2/Si) = loge(a) — bSr, 



Model II: log. (i?, + 2/5;) = loge(c) — dPi+i. 



Estimates of recruitment by the Murphy 

 method varied from a high of 1.5 billion shrimp 

 on May 1, 1962, to a low of 0.6 billion on May 1, 

 1968. Spawning stocks producing recruitment 

 varied from 4.5 million pounds in September 

 1959 to 1.8 million pounds in September 1963 

 (Table 9). 



The range in recruitment observed at any giv- 

 en spawning stock size was very large relative 

 to the range in size of spawning stock, and the 

 fitting of Model I did not result in a meaningful 



1036 



