FISHERY BULLETIN: VOL. 75, NO. 3 



$20 



< 



t- 17.5 



X 



tS 



3 



< 



« 15.0 



< 



Q 



O 12.5 



I- 



H 



X 



g 100 



< 



H 

 r 

 B 



2 



b- 

 O 



O 



I- 

 < 



7.5 



5 



2 5 



I 



R • 0.3470e^ 



,-• 



"f""f I L. 



J 1 I L_ 



1.5 2.5 3.5 4.5 5.5 6.5 8.5 10.5 12.5 4.5 16.5 



MIDPOINT OF LENGTH CLASS (mm) 



FIGURE 9.— Night to day ratios of sums of catches, standardized 

 to numbers under 10 m 2 of sea surface, for thread herring larvae 

 collected in 1971-74 in the eastern Gulf of Mexico. The ratios 

 were calculated for larvae within each 1-mm length class from 

 1.1 to 17.0 mm SL. A fitted exponential regression describes the 

 relationship. Larval abundance estimates for each length class 

 at stations occupied during daylight were corrected by the ap- 

 propriate ratio factor for each length class to account for daytime 

 avoidance. 



herring larvae (Houde 1977a) showed relative in- 

 creases in night catches until larvae were 13.0 

 mm; then the ratio declined to unity for larger 

 larvae. In scaled sardine larvae (Houde 1977b), 

 the ratio increased throughout the larval size 

 range, but the relative increase in night catches 

 was slight compared to thread herring. 



Annual estimates of larval abundance by 1-mm 

 length classes were calculated for 1971 and 1973 

 (Figure 10), after the data had been corrected for 

 daytime avoidance. Abundance of larvae was 

 slightly higher in 1973 than in 1971. The abun- 

 dance of 3.0- to 7.0-mm larvae accounted for the 

 difference between the two years (Figure 10). Lar- 

 vae longer than 17.0 mm were more abundant in 

 1973 than in 1971. 



Abundance of larvae decreased exponentially in 

 both years as lengths increased (Figure 10). Expo- 

 nential functions were fitted to data in the 4.1- to 

 19.0-mm length classes in 1971 and to the 5.1- to 

 20.0-mm length classes in 1973 (Figure 10), giving 

 estimates of the instantaneous decline in abun- 

 dance of thread herring larvae per millimeter in- 

 crease in length. The instantaneous coefficients 

 estimate larval mortality rates if gear avoidance 



140 



120 



IOO 



80 - 



60 



40 - 



20 



Z o 



111 



o |80 



< 

 Q 



Z 



CD 



< 160 



O 



UJ 



t- 

 < 

 X 140 



120 



100 



80 



60 



40 



20 



1971 



I MS (13.7572 x IO ,3 )e 03545L 



r * ■.. ) .. » .. l i.*.. r .ii.Y' * ■. r , » ,, r «, t ,, i «, r .»i r i. F . T .«. r 



,,w; p3 



-  *  ' i « 



1973 



I ^_N L - (17.9238 xlO l3 )e 03942L 



AS" 



I.I- 2.1-3.1- 4.1- 6.1- 8 1- 10 1- 12.1- 14.1- 16 1 18 1- 20.1- 22 1- 

 2 3040 50 7,0 90 110 130 15 170 19 21.0 23 



LENGTH-CLASS (mm) 



FIGURE 10. — Length- frequency distribution of annual larval 

 abundance estimates for thread herring larvae collected in the 

 eastern Gulf of Mexico, 1971 and 1973. Frequencies in each 

 1-mm length class are expressed as estimated annual abundance 

 and have been corrected for daytime avoidance. Fitted exponen- 

 tial functions provide estimates of the instantaneous coefficient 

 of decline in abundance by length. 



is not too great over the length ranges in the 

 analysis. Coefficients wereZ = 0.3545 in 1971 and 

 Z = 0.3942 in 1973. The corresponding percentage 



506 



