HOUDE: ABUNDANCE AND POTENTIAL YIELD OF THREAD HERRING 



1000 



~I00 



b 



< 



Q 



z 



3 

 CD 



< 



Q 



111 



o 



LU 



cc 

 cc 



o 



CJ 

 I 



z 



l- 

 < 

 cc 



3 

 Q 



10 



I - 



I 



1973 

 N t - (865.78 x 10 "le 025 "' 



© 



® 



1971 

 N,- (337.80x10 ") e ° 2l24t 



Opisthonema oglinum survival 

 • • 1971 

 - -197 3 



6 8 10 12 14 16 



ESTIMATED MEAN AGE (DAYS) 



18 20 22 



FIGURE 11. — Estimated abundance of egg and larval stages of 

 thread herring in the eastern Gulf of Mexico in 1971 and 1973. 

 Abundance is expressed as a function of estimated age. Fitted 

 exponential functions give estimates of the instantaneous rates 

 of decline in abundance for eggs and larvae up to 21 days of age. 

 The two symbols enclosed in circles represent nonfully vulner- 

 able length classes and were not included in the regression of 

 instantaneous decline. 



abundance of young larvae was underestimated in 

 each year. If this is true, then mortality estimates 

 (Table 8) are too low. Also, if growth was not expo- 

 nential, but linear, then abundance of larvae in 



older age-classes was overestimated and mortality 

 rates of thread herring larvae would be greater 

 than estimates from the regression coefficients 

 (Table 8). 



Houde ( 1 977a) estimated instantaneous mortal- 

 ity coefficients from abundance at age data for 

 round herring larvae to be Z = 0.1317 in 1971-72 

 and Z = 0.1286 in 1972-73. These estimates are 

 lower than the most probable rates for thread her- 

 ring larvae. The estimated mortality coefficient (Z 

 = 0.2835) for scaled sardine larvae in 1973 was 

 similar to those for thread herring (Houde 1977b). 

 The thread herring instantaneous mortality 

 coefficients for abundance at age data were similar 

 to those for Pacific sardine (Z = 0.16-0.17) 

 (Ahlstrom 1954), jack mackerel (Z = 0.23)(Farris 

 1961), and Japanese mackerel (Z = 0.3295) 

 (Watanabe 1970), but higher than those reported 

 for Japanese sardine (Z = 0.1279) (Nakai and Hat- 

 tori 1961 ) or plaice (Z = 0.0209 to 0.0685) (Bannis- 

 ter et al. 1974). 



Estimated numbers and percentage survival of 

 thread herring at hatching, 5.5 mm SL, and 15.5 

 mm SL were calculated given three possible in- 

 stantaneous growth rates, corresponding to mean 

 daily growth increments of 0.6, 0.8, and 1.0 mm 

 (Table 9). The estimating procedure used 

 parameters from the exponential functions de- 

 scribing decline in numbers by age (Table 8) and 

 the age-at-length data assuming exponential 

 growth (examples in Table 7). The estimated 

 number of spawned eggs, from Table 3, varied by 

 more than a factor of three between 1971 and 

 1973, yet the estimated number of survivors when 

 larvae begin to transform to juveniles (15.5 mm 

 SL) (Richards et al. 1974) was not much different 

 between years (Table 9). Percentage survival from 

 spawned egg to that stage did vary between 1971 

 and 1973; an estimated mortality of >99 c 7c oc- 

 curred in 1973, but mortality was approximately 



TABLE 9. — Estimated numbers and percentages of survivors of thread herring at hatching, 5.5 mm SL, and 15.5 mm SL in 1971 and 

 1973. Estimates are made at three possible growth rates (see Table 8). Duration of the nonfully vulnerable larval stages was set at 1.0 

 days for 1.1 to 4.0 mm larvae in 1971 and at 4.0 days for 1.1 to 5.0 mm larvae in 1973. The number of spawned eggs in each year was 

 based on the estimates in Table 3. Predicted numbers at hatching, 5.5 mm, and 15.5 mm are calculated from exponential functions 

 based on Table 8 data. 



'Hatching assumed to occur at 0.84 days. 



509 



