FISHERY BULLETIN: VOL. 75, NO. 3 



values (0.5, 0.75, and 1.00) within the probable 

 range for thread herring. Based on the mean of 

 1971 and 1973 biomass estimates, potential yield 

 ranged from 60,300 to 120,600 metric tons. It is 

 likely that the sustainable yield of adult stock was 

 in that range during 1971-73. Assuming thread 

 herring are evenly distributed within the 106.7 x 

 10 5 ha spawning area, then probable harvestable 

 yields of adult thread herring range from 5.6 to 

 11.3 kg/ha. Yield could be supplemented by some 

 additional catch of juveniles. 



The eastern Gulf thread herring stock appar- 

 ently is not as large as the menhaden stock in the 

 north-central Gulf. But, a potential harvest, based 

 on 1971-73 biomass levels, of about 100,000 met- 

 ric tons substantiates the belief that thread her- 

 ring are a significant resource in the eastern Gulf 

 that could provide raw material for the fishmeal 

 industry. Because large fluctuations in thread 

 herring year-class strength may occur, yield in 

 some years could be considerably higher than that 

 predicted based on 1971-73 abundance. The po- 

 tential for thread herring harvest is higher in the 

 eastern Gulf of Mexico than that estimated along 

 the Atlantic coast by Pristas and Cheek (1973). 



Larval Abundance 



Larval abundance varied seasonally with peak 

 abundance in spring and summer months (Table 



TABLE 6. — Abundance estimates of thread herring larvae for 

 each cruise. Estimates include larvae in all size classes and were 

 obtained using Equations (2) and (3) (Houde 1977a). 



1 Positive area is defined as the area representing stations where either eggs 

 or larvae of thread herring were collected. 



2 Values are not adjusted for cruises that did not encompass the entire area, 

 nor have estimates been corrected to account for gear avoidance by larvae at 

 stations sampled in daylight. 



6). Abundance estimates for cruises in which 

 thread herring larvae were collected ranged from 

 0.16 to 107.57 x 10 10 larvae in the survey area. 

 Thread herring larvae were collected in small 

 numbers on three cruises in which no eggs were 

 taken (Table 1). Cruises IS7205 and IS7313 were 

 made in late summer when eggs, if present, must 

 have been rare. Larvae collected in early March, 

 during cruise CL7405, occurred only in the south- 

 ernmost part of the survey area (Figure 6). They 

 occurred at five stations on that cruise but abun- 

 dances were only 0.6 to 4.4 under 10 m 2 . The pres- 

 ence of larvae indicated that some spawning 

 began as early as February and that it continued 

 as late as September. 



The seasonal nature of thread herring larvae 

 abundance can be observed in plotted length- 

 frequency distributions for each cruise in which 

 larvae were collected (Figure 8). Larvae were rep- 

 resented in length classes up to 23.0 mm SL, but 

 specimens longer than 15.0 mm were uncommon. 

 The smallest length classes (1.1-3.0 mm) repre- 

 sent larvae in poor condition or that were distorted 

 from net capture and preservation, because re- 

 cently hatched thread herring larvae are 3.8 to 4.0 

 mm SL (Richards et al. 1974). 



Fewer larvae were collected at stations sampled 

 during the day than at night, indicating that gear 

 avoidance was relatively great during daylight, 

 particularly by larger larvae. The ratio of night 

 catches to day catches increased rapidly when 

 summed catches under 10 m 2 over all cruises were 

 plotted for each 1-mm length class (Figure 9). No 

 larvae longer than 17.0 mm were collected during 

 daylight. An exponential function R = 

 0.3470e° 2492X was fitted to the plotted data for lar- 

 vae up to 17.0 mm (Figure 9), where/? is the ratio 

 of night-caught to day-caught larvae and X is 

 standard length. It provided the correction factor 

 R (Houde 1977a), by which daytime catches were 

 adjusted to obtain abundance estimates of larvae 

 by 1-mm length classes in each station area on a 

 cruise. The correction for under-sampling during 

 daylight probably did not completely account for 

 gear avoidance by larvae (Smith and Richardson 

 in press), but it helped to provide a better estimate 

 of larval abundance for subsequent estimation of 

 survival rates. The observed increase in ratio of 

 night- to day-caught thread herring larvae 

 throughout the larval period seems typical of 

 clupeid larvae ( Ahlstrom 1954, 1959; Lenarz 1973; 

 Matsuura in press). But, observations on round 



504 



