60 



Fishery Bulletin 105(1) 



side the closed area. We redistributed effort during the 

 closure to other areas in proportion to historical effort 

 within the same seasonal quarter. Previous studies have 

 chosen not to redistribute effort (Goodyear, 1999), to 

 redistribute effort in proportion to target catch (Worm et 

 al., 2003), or to redistribute effort with the assumption 

 that catch, rather than effort, is a limiting factor (Baum 

 et al., 2003). A good understanding of fleet dynamics is 

 necessary to determine appropriate models for effort 

 redistribution. 



Second, we assumed that redistributed effort would 

 yield the same CPUE as previous effort in the area. 

 Redistributed effort assumes that CPUE will remain 

 unchanged when more fish are removed. It is likely that 

 CPUE would decline with abundance as a result of in- 

 creased effort, therefore it is possible that our analysis 

 overestimates the catches during the closure. Similarly, 

 our model assumes that CPUE is constant within each 

 area, i.e. regardless of where one fishes within the area, 

 one achieves the same CPUE. In reality, it is possible 

 that fishermen could fish close to the edge of the closed 

 area and potentially undermine the effectiveness of a 

 closure. 



Related to these first two points is the case of switch- 

 ing between fishing modes. By grouping FOB and UNA 

 sets in our model, we allowed for switching between set 

 types when fishing outside the closed area. Harley et 

 al."* showed that the purse-seine vessels that catch the 

 majority of the bigeye tuna, fish almost exclusively on 

 floating-objects (over 90% of the sets). Even with this 

 information, we still believe that the implicit assump- 

 tion of grouping the two set types is acceptable. We did 

 not consider dolphin-associated sets (that catch almost 

 exclusively yellowfin tuna). We consider it much less 

 likely that effort would be shifted towards dolphin-as- 

 sociated schools for several factors, including politics, 

 market pressure, technological and gear differences, 

 and the inexperience that many skippers who partici- 

 pate in the FOB fishery would have with this alterna- 

 tive mode of fishing. 



Finally, we implicitly assumed in our model that fish 

 not caught as a result of the closure could not be caught 

 later in the year. This assumption could lead us to 

 underestimate catches outside of the closure. Thus, we 

 have two potential biases in opposite directions that 

 could affect our conclusions. The best way to quan- 

 tify these biases would involve a model that integrated 

 population and fisheries dynamics. 



A dynamic approach to modeling closed areas could 

 take into account the abundance of fish in different 

 areas and the movement of fish between areas dur- 

 ing the year. Modeling the relationship between effort 

 and catches in different areas should include account- 

 ing for abundance (e.g., through the use of the catch 

 equation). 



Tagging data are necessary to estimate stock param- 

 eters, such as residence times within a closed area and 

 fish movement rates between the open and closed areas. 

 In addition to conventional tagging data, information 

 from electronic tagging of bigeye tuna (Schaefer and 



Fuller, 2002) could provide a basis for describing move- 

 ment by means of simple movement models (e.g., those 

 of Adam et al. [2003]). Because the vessels catch bigeye 

 and skipjack tunas together, the model must include the 

 movement patterns of both species. 



This approach is extremely data demanding, and 

 many of the data for this approach are not yet avail- 

 able. Notwithstanding these problems, future analysis 

 of time-area closures should include consideration of im- 

 portant biological factors such as those described above, 

 as well as socioeconomic data that may be important for 

 predicting fleet dynamics. 



Another extension of the modeling approach in our 

 study is to consider additional target and bycatch spe- 

 cies. Worm et al. (2003) considered bycatch from the 

 United States swordfish and tuna longline fisheries 

 in the Atlantic when modeling closed areas. With this 

 approach it would be useful to include not only yel- 

 lowfin tuna and dolphin sets in the model, but also the 

 bycatch species that are taken in the different areas 

 and fisheries. 



Conclusions 



Time-area closures are one of the many management 

 actions available for the regulation of fisheries. Because 

 of the strong interactions between bigeye and skipjack 

 tunas, we have shown that time-area closures alone are 

 unlikely to be sufficient to address concerns regard- 

 ing the sustainability of bigeye tuna because it may 

 not be possible to achieve the necessary reductions in 

 bigeye tuna catches without large losses in skipjack tuna 

 catches. We suggest that it will be important to investi- 

 gate aspects offish behavior to determine measures that 

 could be used either in conjunction with, or instead of, 

 closures to help reduce mortality on juvenile bigeye tuna 

 while sustaining the important skipjack fishery. 



Acknowledgments 



We thank R. Allen, W. Bayliff, R. Deriso, and three anon- 

 ymous reviewers for comments on this manuscript, and 

 M. Hall, C. Lennert-Cody, M. Maunder, and K. Schaefer 

 for useful discussions of closed areas and management 

 options for bigeye tuna. 



Literature cited 



Adam, M. S., J. Sibert, and D. Itano. 



2003. Dynamics of bigeye [Thunnus obesus) and yel- 

 lowfin (T. albacares) tuna in Hawaii's pelagic fisheries: 

 analysis of tagging data with a bulk transfer model 

 incorporating size-specific attrition. Fish. Bull. 101: 

 215-228. 

 Baum, J. K., D. Kehler, R. A. Myers, B. Worm, S. J. Harley, and 

 P. A. Doherty. 



2003. Collapse and conservation of shark populations in 

 the Northwest Atlantic. Science 299:389-392. 



