138 



Fishery Bulletin 101(1) 



based on a discrete time origin that was centered in the 

 spawning season and it was assumed that samples were 

 obtained at that time. 



For simpHcity, constant annual recruitment to the pop- 

 ulation at age 1 is assumed, although this particular 

 assumption is later relaxed and its specific effect on es- 

 timates of population weight-specific fecundity is evalu- 

 ated. To start the simulated population, recruitment 

 was arbitrarily set equal to N, = 1.0/yr for both females 

 and males. Then, given female weight at age (Fig. 3) 

 and female numbers at age (Fig. 4), one can calculate 

 age-specific female cohort biomass as the product of 

 numbers and individual weights (Table 3), which when 

 summed over all ages yields the total equilibrium female 

 biomass (411.37 g/female recruit). Likewise, age-spe- 

 cific cohort fecundity is calculated a.s female numbers 

 at age, multiplied by estimates of individual female fe- 

 cundity (Fig. 5), multiplied by the proportion of females 



that are mature (Fig. 6), which when summed over all 

 ages classes yields the expected lifetime larval produc- 

 tion of a cohort (42,028 lai^vae). The ratio of these two 

 quantities (0=102.17 larvae/g of female) estimates the 

 population weight-specific fecundity of female shortbelly 

 rockfish. This population statistic can be compared with 

 individual age-specific estimates presented in Table 3. 

 These weight-specific fecundities range from 77.90 lar\'ae/g 

 offemale for the mature 1-yr-old females, to 114.98 larvae/g 

 of female for the oldest fish. 



It is also revealing to compare predicted weight-specific 

 fecundity at age from the life table analysis (Table 3i with 

 the observed data calculated directly from individual fish 

 (Fig. 7). Results show a great deal of variability in the 

 observed data, which is consistent with the extensive 

 residual variability in fecundity (see Fig. 5, Table 2). Life 

 table i)re(iictions were generally similar to, but slightly 

 higher than, the observed data. This slight difference is 



