McGOWAN and RICHARDS: BLUEFIN TUNA LARVAE 



lations of potential stock size of spawners off the 

 southeastern U.S. 



Additional data indicate that there is only one 

 spawning stock in the northwestern Atlantic. 

 Bluefin tuna are known to occur in different 

 places at different times of the year, depending 

 on size and age (Rivas 1978; Mather 1980). The 

 large adults are expected to be migrating 

 through the Straits of Florida during late April 

 and May after spawning in the Gulf of Me.xico. A 

 few migrating, ripe females and recently spent 

 males were caught in May near Bimini in the 

 Bahamas (Rivas 1954). Bluefin tuna are capable 

 of swimming from the Bahamas to Norway at 

 sustained speeds as fast as 122 miles per day 

 (Brunenmeister 1980), so adult fish could easily 

 traverse the area from Miami to Cape Hatteras, 

 or be widely distributed over the Blake Plateau, 

 a few days after leaving the Gulf of Mexico. They 

 could migrate back and forth between the two 

 areas during the 60 d spawning season, although 

 there is no evidence for this. Because there is no 

 evidence for two separate groups of spawning 

 fish, the parsimonious assumption is that there is 

 only one. Therefore we assumed that fish in both 

 areas had the same reproductive parameters 

 previously estimated (Baglin 1982; McGowan 

 and Richards 1986). 



The estimated age-at-length of bluefin tuna 

 larvae was based on previous analysis of daily 

 increments in otoliths of larval bluefin collected 

 from the Gulf Stream near Miami (Brothers et 

 al. 1983). We calculated a linear regression to 

 predict age from length using the mean esti- 

 mates of age at length presented in Brothers et 

 al. (1983). The equation is 



Age (days) = 3.67 x standard length (mm) 

 - 8.04. 



This equation was based on limited ranges of 

 age and length, so we used it heuristically as the 

 best available. It may be revised after further 

 study extends the age and length data, but the 

 revisions will most hkely be at the older-longer 

 end of the relationship, not at the younger- 

 shorter end most relevant to our conclusions in 

 this paper. There is a range of age at length 

 which could affect interpretations of time spent 

 drifting by the larvae but our use of the mean 

 results in conservative interpretations of the 

 data in most instances. 



In this paper we refer to the current from the 

 Dry Tortugas to Cape Hatteras as the Gulf 

 Stream (Iselin 1936; Stommel 1965). We refer to 



the continental shelf area between Palm Beach, 

 FL and Cape Hatteras, NC as either the region 

 off the southeastern United States, or the South 

 Atlantic Bight. 



Stations occupied during RV Oregon II cruise 

 152 were numbered 42XXX, where XXX is a 

 sequential station number. For brevity we refer 

 to stations in this paper by their unique 3 digit 

 number, the XXX part. 



RESULTS 



Catch and Abundance of Larvae 



Larval bluefin tuna were collected at 10 of 147 

 stations during cruise 152 (Fig. 1). Three larvae 

 were collected at one station, two at two sta- 

 tions, and one each at the other positive stations 

 for a total catch of 14 larvae (Table 1). To put this 

 small catch in perspective, in 1984 and 1986 the 

 average catch in the Gulf of Mexico was less than 

 24 total larvae at 10 positive stations (McGowan 

 and Richards 1987). Thus the 14 caught in 1985 

 could have been over 50% of the expected catch 

 for the Gulf of Me.xico in 1985. The larvae ranged 

 in length from 3.0 to 6.2 mm corresponding in 

 age from 3 to 14.7 days postfertilization. The 

 estimated mean abundance of bluefin tuna larvae 

 from stations at or outside the 183 m isobath was 

 0.383 ± 0.114 (SE) under 10 m" of sea surface 

 (approximately V.i the density of bluefin larvae in 

 the Gulf of Mexico). The corresponding area sur- 

 veyed was 2.02 X 10^^ m" producing a total esti- 

 mated 7.74 X 10^ larvae in the survey area. 

 These larvae could have been produced by 3,730 

 adult fish weighing a total of 903 t. This is equiv- 

 alent to about 5% of the 1985 estimate of Gulf of 

 Mexico spawning stock calculated from the 

 larval index (McGowan and Richards 1987). The 

 coefficient of variation of the estimate of abun- 

 dance of these larvae was 30%, which is in the 

 range of coefficients of variation for the Gulf of 

 Mexico for the past 10 years, 21-49% (McGowan 

 and Richards 1987). 



Distribution of Bluefin Tuna Larvae and 

 Coincident Water Masses 



There were three groups of stations where 

 bluefin tuna larvae were present: 1) three sta- 

 tions at the shelf break east of Florida, 2) two 

 stations near the shelf break off North Carolina, 

 and 3) the positive stations over the Blake 

 Plateau. The stations in the first group (634, 636, 

 and 647) were near the 183 m isobath. Two inde- 



617 



