140 



Use of genetic data to assess 

 the uncertainty in stock assessments 

 due to the assumed stock structure: 

 the case of albacore {Thunnus alalungd) 

 from the Atlantic Ocean 



Haritz Arrizabalaga^ (contact author) 



Victoria Lopez-Rodas^ 



Eduardo Costas^ 



Alberto Gonzalez-Garces^ 



Email address for H. Arrizabalaga: harri@pas.azti.es 



' AZTI Tecnaha 

 Herrera Kaia 

 Portualdea z/g 

 20110 Pasaia (Gipuzkoa) Spam 



2 Universidad Complutense 

 Genetica, Facultad de Vetennaria 

 Avda. Puerta de Hierro s/n 

 28080 Madrid. Spain 



3 lEO 



Cabo Estay, Canido 



Apdo 1552 



Vigo (Pontevedra), Spain 



Stock assessments can be problematic 

 because of uncertainties associated 

 with the data or because of simpli- 

 fied assumptions made when model- 

 ing biological processes (Rosenberg 

 and Restrepo, 1995). For example, 

 the common assumption in stock 

 assessments that stocks are homo- 

 geneous and discrete (i.e., there is no 

 migration between the stocks) is not 

 necessarily true (Kell et al., 2004a, 

 2004b). 



On the other hand, it is essential 

 that the stock structure assumed 

 during the assessment and manage- 

 ment process corresponds to the real 

 population structure of the resource. 

 Otherwise, fishery management be- 

 comes inefficient (less productive 

 populations may be overfished and 

 collapse, while more productive popu- 

 lations may be underexploited [Allen- 

 dorf et al., 1987; Begg et al., 1999]) 

 and may affect biological attributes, 

 such as growth, productivity, or ge- 

 netic diversity (Ricker, 1981). In spite 

 of this problem, current regulations 

 on several fisheries are based on spa- 

 tial schemes that do not necessarily 

 reflect the real biological structure 



of the populations (Pawson and Jen- 

 nings, 1996; Stephenson, 1999; Ward, 

 2000). In these cases, the results of 

 stock assessments may be biased 

 and, in general, an important level 

 of uncertainty exists in stock assess- 

 ments (NEC, 1994; Turner, 1998) due 

 to the assumed stock structure. 



An assessment of the magnitude 

 of this uncertainty is important so 

 as to increase confidence in the as- 

 sessment itself. Moreover, quantify- 

 ing the uncertainty allows the evalu- 

 ation of the relative effect of stock 

 structure assumptions with respect 

 to other assumptions about biologi- 

 cal, fishery, or modeling parameters 

 in the assessment. Knowing the rela- 

 tive importance of the effect of these 

 underlying assumptions will allow 

 management scientists to prioritize 

 the types of research needed to bet- 

 ter ground the stock assessments 

 with real information. 



In this note, we suggest a way to 

 assess uncertainty in stock assess- 

 ments that is due to assumptions 

 of stock structure. The assessment 

 is essentially based on a sensitivity 

 analysis conducted by testing alter- 



native stock structure hypotheses 

 derived from available genetic, fish- 

 ery, and biological information. The 

 method is illustrated with albacore 

 (Thunnus alalunga, Bonn. 1788) in 

 the Atlantic Ocean. 



Albacore is a highly migratory spe- 

 cies distributed between latitudes 

 45°N and 45°S. Studies of albacore 

 reproduction in the Atlantic Ocean 

 have shown different spawning pe- 

 riods and areas in both hemispheres 

 (Beardsley, 1969; Koto, 1969). Shio- 

 hama (1971) and Uozumi (1996), 

 based on Japanese longline distribu- 

 tion studies, described an adult con- 

 centration area in each hemisphere. 

 These findings, along with studies of 

 larval concentration areas (Ueyanagi, 

 1971), support the existence of two 

 separate populations, one in each 

 hemisphere. Based on these studies, 

 it is assumed within the Interna- 

 tional Commission for the Conserva- 

 tion of Atlantic Tunas (ICCAT) that 

 there are two albacore management 

 units in the Atlantic, separated by 

 parallel 5°N. However, various au- 

 thors have suggested the possibil- 

 ity that albacore move between the 

 north and south Atlantic (reviewed 

 in Gonzalez-Garces, 1997). Moreover, 

 the continuous spatial distribution of 

 catches around the equator also sug- 

 gests this possibility (Fig. 1). 



Recent studies have shown ge- 

 netic differences between north and 

 south Atlantic albacore (Takagi et 

 al., 2001; Arrizabalaga et al., 2004), 

 but it is still unclear whether the 

 limit between both populations is at 

 latitude 5°N or somewhere else. In 

 fact, results from Arrizabalaga et 

 al. (2004) are not concordant with 

 the limit at latitude 5°N because a 

 sample from the Gulf of Guinea (1°N, 

 15-16°W) was genetically more like 

 the sample from the north Atlantic 

 than the one from the south Atlan- 

 tic. This observation may indicate 

 that either the limit between both 

 stocks may be located farther south 

 than that currently assumed or that 



Manuscript submitted 15 June 2005 

 to the Scientific Editor's Office. 



Manuscript approved for publication 



14 December 2005 by the Scientific Editor. 



Fish. 105:140-146(2007). 



