Nemerson et al.: Spawning site fidelity in Thunnus thynnus 



119 



These two extremes, and the terms used to describe 

 them, have been the subject of some confusion in the 

 bluefin tuna hterature. The permanent transfer of 

 individuals from one side of the Atlantic to the other 

 has been called the "no-memory condition" (Punt and 

 Butterworth, 1995; Powers and Cramer, 1996) or the 

 "diffusion model" (Cooke and Lankester, 1996). The 

 case where migrants always return to their natal 

 side prior to the next spawning season is termed the 

 "overlap model" by Cooke and Lankester (1996). We 

 follow the convention of Cooke and Lankester (1996) 

 and use the terms diffusion and overlap to refer to the 

 two models. We use "transfer rate" to refer to the per- 

 manent transfer of an emigrant from one population 

 to the other and "migration rate" to refer generally to 

 the trans-Atlantic movement of individuals. Finally, 

 we use the term "memory" to refer only to an individ- 

 ual's behavior with respect to spawning location, not 

 to other life history attributes. That is, under the dif- 

 fusion (no-memory) model, a migrant will spawn on 

 the side of the ocean to which it migrates, regardless 

 of its birth location, but will retain other life history 

 attributes such as size or age at maturity. 



The permanent transfer of individuals can be con- 

 sidered a migration for dispersal (Greenwood and 

 Harvey, 1982), whereas the overlap model can be 

 assumed to be a feeding migration, and is free of 

 implications for reproductive mixing. One can envi- 

 sion intermediate scenarios combining varying degrees 

 of memory, or philopatry. For example, migrants may 

 remain on the opposite side for a period of years, 

 while either participating in or foregoing spawning, 

 before ultimately returning to their natal side. Fur- 

 thermore, some migrants may exhibit spawning 

 site fidelity while others may stray, joining previ- 

 ously established spawning populations (e.g. Curry, 

 1994). 



Simulation models have shown that the dynamics 

 of the two populations are potentially very sensitive 

 to even low trans-Atlantic migration rates, partic- 

 ularly for east-to-west transfer (NRC, 1994; Porch 

 et al., 1995; Punt and Butterworth, 1995; Powers 

 and Cramer, 1996) because the average size of the 

 eastern population has been about 6 to 13 times 

 that of the western population over the past 20 

 years (ICCAT^'^; Fig. 1). Recent spawning biomass 

 estimates for the western population are based on 



' ICCAT ( International Commission for the Conservation of Atlan- 

 tic Tunas). 1994a. WestAtlanticbluefin tuna. Biennial report 

 of the ICCAT Standing Committee on Research and Statistics, 

 41 p. ICCAT, Estebanez Calderon 3, E-28020, Madrid. Spain. 



^ ICCAT (International Commission for the Conservation of Atlan- 

 tic Tunas). 1994b. East Atlantic bluefin tuna. Biennial report 

 of the ICCAT Standmg Committee on Research and Statistics, 

 31 p. ICCAT, Estebanez Calderon 3, E-28020, Madrid, Spain. 



East Atlantic 

 Population 



West Atlantic 

 Population 



Figure 1 



Representation of the effect of migration on the eastern 

 and western populations of bluefin tuna. Migration from 

 the larger eastern population to the west has a larger effect 

 on the western population than does migration from the 

 smaller western population to the east. In this schematic, 

 the eastern population is about six times the size of the 

 western population, and the migration rates are about 1% 

 in each direction. 



catches throughout the fishing area, which includes 

 the entire North Atlantic west of 45°W longitude. 

 If fish of eastern origin are included in these catch 

 statistics but do not spawn in the west Atlantic, 

 then western spawning biomass will be substantially 

 overestimated (Powers and Cramer, 1996; American 

 Fisheries Society'^). 



Determining the spawning site fidelity of itero- 

 parous pelagic species that occur over a wide area of 

 open ocean is difficult. Population differentiation can 

 be inferred from tag-return data, comparisons of life 

 history parameters and morphometric characters, 

 or from genotypic variation. Several studies have 

 attempted to analyze the population structure of 

 Atlantic bluefin tuna with these methods (Calaprice, 

 1986; NRC, 1994; Cooke and Lankester, 1996). 



Several investigators have reviewed and ana- 

 lyzed trans-Atlantic tag returns to estimate rates 

 of migration (NRC, 1994; Punt and Butterworth, 

 1995; Turner and Powers, 1995; Cooke and Lank- 

 ester, 1996). These studies have estimated annual 

 migration rates of between 1% and 10% and have 

 considered both diffusion and overlap models. Gen- 

 erally, these studies have sought to find interpreta- 

 tions of tag-return data that agree best with other 

 estimates of population size. 



•* American Fisheries Society. 1995. Marine Fisheries Section 

 statement on bluefin tuna, 2 p. Am. Fish. Soc, 5410 Grosvenor 

 Lane, Ste. 110, Bethesda, MD 20814-2199. 



