HOUDE: ABUNDANCE AND POTENTIAL YIELD OF THREAD HERRING 



Richardson (in press). Variance estimates on 

 cruise and on annual egg abundance estimates 

 were calculated by methods used by Cushing 

 (1957) and Taft (1960). Houde (1977a) has given 

 detailed procedures, including estimating for- 

 mulae, that were used to obtain abundance esti- 

 mates of clupeid eggs and larvae in eastern Gulf of 

 Mexico surveys. 



Two methods were used to estimate adult 

 biomass, based on two different procedures for de- 

 termining annual spawning by thread herring. 

 The first procedure is that given by Sette and 

 Ahlstrom (1948). The estimate of annual spawn- 

 ing depends on integrating station and cruise es- 

 timates over area and time. The second procedure 

 is based on a modification of Simpson's (1959) 

 method in which annual spawning is estimated by 

 plotting the daily spawning estimates for each 

 cruise against the middate of the cruise and then 

 determining the area under the resulting polygon 

 by planimetry. 



Potential Yield to a Fishery 



Houde (1977a) used the estimator suggested by 

 Alverson and Pereyra (1969) and Gulland (1971, 

 1972) to predict potential yield of round herring in 

 the eastern Gulf. The same procedure was used for 

 thread herring. The estimating formula is C max = 

 XMB where X is assumed to equal 0.5, M is the 

 natural mortality coefficient, and Bo is the virgin 

 biomass. My biomass estimates are estimates ofB 

 since the thread herring stock is virtually 

 unfished in the eastern Gulf. Because no estimate 

 of M exists for thread herring, the potential an- 

 nual yield was predicted using a range of probable 

 values of the mortality coefficient. 



Larval Abundance and Mortality 



Mortality estimates were determined for larvae 

 by length and by estimated ages. The exponential 

 decrease in abundance of 1-mm length classes was 

 used to calculate mortality coefficients to describe 

 the decline in catches by length. Growth was as- 

 sumed to be exponential during the larval phase. 

 Based on this assumption and information on 

 laboratory growth rates for thread herring larvae, 

 ages of larvae in 1-mm length classes were esti- 

 mated. Mortality coefficients were then estimated 

 from the decline in abundance of larvae in relation 

 to estimated age. Houde (1977a) gave estimating 

 formulae and discussed the rationale for his pro- 



cedures, which are similar to those used previ- 

 ously by Ahlstrom (1954) and Nakai and Hattori 

 (1962). 



RESULTS AND DISCUSSION 



Occurrence of Eggs and Larvae 



Thread herring eggs occurred in 8 of the 17 

 cruises from 1971 to 1974, and larvae occurred 

 during 11 of the cruises (Table 1). Eggs were col- 

 lected on cruises from May through August, al- 

 though significant spawning may have occurred 

 during April when no cruises were scheduled. 

 Some larvae were collected as early as March and 

 as late as September, but they were most abun- 

 dant from May through August. No eggs or larvae 

 were collected from September through January. 

 Fuss et al. (1969) reported ripe or nearly ripe adult 

 thread herring from the eastern Gulf in March 

 through August. My data support their finding 

 that thread herring spawning is confined to spring 

 and summer in this area. 



Most spawning takes place within 50 km of 

 shore on the inner continental shelf in depths <30 

 m, and virtually all spawning occurs within 100 

 km of shore at depths <50 m (Figure 2). A single 

 instance of egg occurrence beyond the 50-m depth 

 contour was recorded (Figure 2). Spawning was 

 most intense between lat. 26°00'N and 28°00'N, 

 the area from just south of Fort Myers to Tampa 

 Bay, Fla. This is the area where an attempt was 

 made to establish a commercial fishery for thread 

 herring in the 1960's (Fuss 1968; Fuss et al. 1969). 

 Kinnear and Fuss (1971) found that thread her- 

 ring that were concentrated near Fort Myers (lat. 

 26°00'N) in winter migrated north during warmer 

 months. My egg distribution data suggest that a 

 large part of the thread herring population re- 

 mains within the Fort Myers-Tampa Bay area 

 throughout the year. 



Larval distribution was more widespread than 

 that of eggs, presumably due to dispersal by water 

 currents, but was generally similar to egg dis- 

 tribution (Figure 2). Most larvae were collected 

 where water depths were <50 m and only six oc- 

 currences were recorded where depths were >50 m 

 (Figures 2-6). 



Thread herring eggs and larvae were relatively 

 common in eastern Gulf ichthyoplankton. A total 

 of 4,236 thread herring eggs were collected during 

 the 17 cruises, 1.39% of the 304,507 total fish eggs 



495 



