f^ISHERY BULLETIN; VOL. 87, NO. 3, 1989 



Plateau as a spawning area for bluefin tuna. We 

 propose a hypothesis, based on existing data, 

 that bluefin tuna are dependent on a dynamic 

 larval retention area associated with the Loop 

 Current in the eastern Gulf of Mexico. 



The total number of specimens of bluefin tuna 

 larvae on which this paper is based is small. 

 Therefore, we present our conclusions as hypo- 

 theses that are consistent with our data from 

 independent sources and with all other data 

 known to us. Arguments supporting our hypoth- 

 eses and their limitations will be elaborated in 

 the discussion. 



Our investigations were initiated by the obser- 

 vation of bluefin tuna larvae outside the normal 

 spawning area and then proceeded by a series of 

 questions, critical examination of available infor- 

 mation, and conclusions and hypotheses for 

 further investigation. The bluefin tuna larvae off 

 the southeastern U.S. in 1985 could have been 

 spawned locally or transported in currents. 

 Were they in water masses characteristic of the 

 Gulf Stream? If they were in the Gulf Stream, 

 then were they young enough to have been 

 spawned locally or were they transported from 

 upstream? Other researchers had reported inci- 

 dental catches of bluefin tuna larvae in some of 

 the same areas. Were these larvae likely to have 

 been spawned locally? Wherever they were 

 spawned, could they survive where they were 

 collected? What were the general temperature- 

 salinity conditions where bluefin tuna larvae 

 were found off the southeastern U.S.? Are these 

 conditions similar to those which bluefin tuna 

 larvae experience in the Gulf of Mexico? What 

 else is known about the oceanography of this 

 region which is relevant to survival of fish 

 larvae? Given our conclusions from these data 

 and our knowledge of the life history of the blue- 

 fin tuna, what insights can be drawn from the 

 occurrence of larvae outside the presumed 

 spawning area and what hypotheses need to be 

 tested by additional work? 



METHODS 



Ichthyoplankton were collected on cruise 152 

 of the RV Oregon II in April and May 1985. 

 Double obhque tows to 200 m or to near the 

 bottom in shallow water were made with 0.6 m 

 diameter bongo nets having 0.333 mm mesh 

 size. Bluefin tuna larvae were identified and 

 measured by W. J. Richards. Salinity and tem- 

 perature data collected by CTD, XBT, or bottle 

 cast at each station were extracted from com- 



puter data files of the National Marine Fish- 

 eries Service, Mississippi Laboratory, Pasca- 

 goula. Satellite data for April and May 1985 

 were obtained from NOAA Miami SFSS Gulf 

 Stream Position Flow Chart #2450 for the days 

 during the cruise. Historical observations of 

 bluefin tuna larvae and coincident temperature 

 and salinity near Cape Hatteras were obtained 

 from Berrien et al. (1978) and Clark et al. 

 (1969). Previous observations of bluefin tuna 

 larvae outside the Gulf of Mexico were re- 

 viewed for other evidence of spawning in the 

 Straits of Florida or elsewhere (Richards 1976; 

 Brothers et al. 1983). 



Statistical estimates of standardized abun- 

 dance of larvae were made using the delta-distri- 

 bution, an efficient estimation technique when 

 zero counts are observed (Pennington 1983). 

 This is the same method routinely used to con- 

 struct the fishery-independent index of the 

 abundance of Gulf of Mexico spawners. Because 

 of logistical and statistical sampling problems, 

 the estimates of abundance are best regarded as 

 indices calculated in a consistent way and valid 

 for comparative purposes. Unless there is spatial 

 periodicity or patchiness at the same scale as our 

 sampling gi'id, the systematic survey does pro- 

 vide, however, an accurate estimate of mean 

 abundance and its variance (Poole 1974; Ripley 

 1981). The details of the method are provided in 

 McGowan and Richards (1986). The estimate 

 assumed that fecundity, sex ratio, spawning 

 season, and length-weight-age ratios were the 

 same off the southeastern U.S. as in the Gulf of 

 Mexico (Baglin 1982). This assumption was sup- 

 ported by the similarity of length-frequency dis- 

 tributions of incidental catches of adults in the 

 two areas during May 1985. Because our 

 assumptions are important to subsequent argu- 

 ments, the evidence substantiating our reason- 

 ing is given in detail below. 



Approximately 90% of incidentally caught 

 adult bluefin tuna in the gulf and off the south- 

 eastern U.S. during May 1985 were large, 

 spawning-sized fish >190 cm (data provided by 

 S. Turner, National Marine Fisheries Service, 

 Miami Laboratory). There was no statistical dif- 

 ference in proportion of adult spawners between 

 the two areas (Chi-Square = 0.0176; P = 0.89; 

 McGowan and Richards 1987). Thus the avail- 

 able catch data were consistent with our assump- 

 tion that the fish in both areas were similar in 

 terms of size-related reproductive capacity. This 

 was the primary justification for using reproduc- 

 tive parameters of Gulf of Mexico tuna for calcu- 



616 



