FISHERY BULLETIN; VOL. 87, NO. 3. 1989 



dance are made only for comparative purposes. 

 A recent independent review of our techniques 

 concluded that the index of larval abundance 

 probably reflects trends in abundance of adults 

 accurately but that our ad hoc estimate of adults 

 from the larvae resembles VPA estimates coin- 

 cidentally because of our choice of a larval mor- 

 tality rate. Our ongoing research is aimed at 

 improving the precision and accuracy of 

 estimates of abundance of bluefin tuna larvae by 

 combining our increasing knowledge of their 

 biology with improved sampling gear and 

 methods. 



Nevertheless, we are confident that our inter- 

 pretation of the preceding data is justified and 

 reasonable because of decades of accumulated 

 experience with this species, because of the vari- 

 ety of independent sources of data which are 

 consistent with our interpretations, and princi- 

 pally because the most important conclusions 

 and hypotheses presented here are not depen- 

 dent upon quantitative estimates of abundance, 

 but upon the relationship between presence and 

 absence of bluefin tuna larvae with specific 

 oceanographic variables. Despite our confidence 

 in our presentation of the results we readily 

 admit that additional data could falsify our con- 

 clusions. 



Assessment of Spawning off the 

 Southeastern United States 



Assuming that all the larvae caught near the 

 shelf and over the Blake Plateau were spawned 

 near where they were collected, spawning in the 

 area was only a small fraction (5%) of estimated 

 total spawning by the western Atlantic stock 

 (McGowan and Richards 1987). This is similar to 

 a previous estimate (6%) of the number of ripe 

 females passing the Bahamas during May (Rivas 

 1954:310). However, because of the currents in 

 this region, it is not certain that the larvae were 

 spawned near the area where they were col- 

 lected. 



Currents along the outer shelf off the south- 

 eastern U.S. average 0.5-1.0 m s~^ during 

 the summer (Atkinson and Menzel 1985). Gulf 

 Stream surface currents off Florida based on 

 ship drift observations during April-June 

 average 1.5 m s~^ or less (Fuglister 1951). 

 During April-June 1985 the mean northward 

 current, measured at 29°N and at 30 m depth 

 over the 75 m isobath, was 0.53 m s~^ and the 

 maximum current was 1.55 m s~\ approxi- 

 mately 3 kn (Lee et al. 1986). Current velocity at 



the wind-affected surface and farther offshore in 

 the fastest moving region of the Gulf stream 

 would be higher. 



If we use the long-term average of 1.5 m 

 s~^ as the mean northward speed of the sur- 

 face current, then a planktonic fish larva would 

 travel 116 km in one day. The mean distance 

 travelled in 7 days would equal 812 km which is 

 less than the straight-line distance from Cape 

 Hatteras, NC (35°N, 76°W) to Palm Beach, FL 

 (27°N, 80°W), approximately 970 km. These cal- 

 culations suggest that larvae <7 days after fer- 

 tilization, which were found as far north as Cape 

 Hatteras, could have been spawned north of the 

 Straits of Florida. Those more than 7 days old, 

 or larger than about 4 mm (minimum size at this 

 age, not mean; see Brothers et al. 1983), could 

 have been spawned in the Straits of Florida or 

 the southeastern Gulf of Mexico. Ten of the 14 

 larvae collected in 1985 were 4.0 mm or longer, 

 and most of the larvae were caught far south of 

 Cape Hatteras. Therefore, based on current 

 velocity and estimated ages, most of the bluefin 

 tuna larvae were probably not spawned off the 

 southeastern U.S. near the area where they 

 were collected. 



The 3 d old (3.0 mm) bluefin tuna larva col- 

 lected near lat. 34°N could not have been 

 spawned in the Gulf of Mexico because, even at 

 2.5 m s~\ it would have travelled only 667 

 km in three days. A 667 km straight line from 

 where the larva was caught would end at about 

 28°N, which is north of Palm Beach. However, 

 the interpretation must be different for the 

 larger, older larvae. Except for the single 3 mm 

 larva, those in the high velocity core of the Gulf 

 Stream (at the shelf edge) could have been ad- 

 vected from a distance to the south in only a few 

 days. At only 1.0 m 5 ' (a little faster than 2 

 kn) larvae 4-5 mm long, corresponding to ap- 

 proximately 8 days old, could have been ad- 

 vected from off Miami to about 31°N in 5 days. 

 The larvae collected along the shelf edge and 

 over the Blake Blateau in 1985 were in this size 

 range or a little longer. Therefore all but the 3 

 mm larva could have been spawned in the south- 

 eastern Gulf of Mexico or between the Florida 

 Keys and Cuba. Slow gi-owth due to cold-water 

 temperature could explain small larvae far from 

 their spawning area, but in this case the Gulf 

 Stream water was warmer than the water where 

 bluefin tuna larvae were found in the Gulf of 

 Mexico. 



It should be noted that the current circulation 

 east and north of the Bahamas is complicated 



626 



