502 



Fishery Bulletin 93(3), 1995 



Because we did not have direct altimetry capabili- 

 ties, size classes of bluefin tuna photographed in this 

 survey could be generally documented only as large- 

 medium or giants from spotter estimates alone. Spot- 

 ters undoubtedly photographed, and we subsequently 

 counted, some bluefin tuna below commercial size 

 classes (<70 inches SFL). Catch records indicated 

 that lengths of individuals in a school may vary by 

 several inches, but only 10% of a seine catch is al- 

 lowed to be undersized. 4 Because a spotter's survival 

 in the commercial fishing industry depends upon size 

 judgments being made before seine boats expend ef- 

 fort to capture schools, there is a strong selection for 

 accuracy (Williams, 1981; Squire, 1993). 



Despite confidence in spotter estimates (Squire, 

 1993), adequate documentation and validation of 

 their ability to judge size or biomass are lacking. 

 Research to define the accuracy of New England blue- 

 fin tuna spotter estimates, or to explore alternative 

 methods of establishing lengths of photographed 

 bluefin tuna, are clearly needed. Future surveys must 

 obtain more specific information on size classes of 

 photographed fish in order to be used as a point of 

 reference for present CPUE-based models that use 

 total length to assign year class (Anonymous, 1986). 



The majority of bluefin tuna schools were photo- 

 graphed in five areas traditionally fished for giant 

 tuna, including Great South Channel, Wilkinson 

 Basin, Piatt's Bank, and Jeffreys Ledge. This group 

 of areas may reflect the past experience of the spot- 

 ters and their unwillingness to search where fish are 

 not usually found; it may also indicate that giant and 

 large-medium-sized tuna exhibit clumped distribu- 

 tions in New England waters, where oceanic frontal 

 systems, bottom topography, and concentration of 



4 Foster, K. Gloucester Laboratory. Natl. Mar. Fish. Serv., NOAA, 

 30 Emerson Ave., Gloucester, MA 01930. Personal commun., 

 March 1994. 



prey provide favorable feeding and thermal conditions 

 (Laurs et al., 1984; Maul et al., 1984; Roffer, 1987). 



Clumped distributions make redundant counts an 

 underlying problem for aerial assessments. Lacking 

 GPS capabilities in 1993, we acknowledge that we may 

 have counted bluefin tuna schools more than once be- 

 cause we had insufficient data to precisely locate all 

 photographed schools. We have learned, however, that 

 a given spotter is unlikely to photograph the same 

 school twice over a period of a few hours. Once a spot- 

 ter directs a boat onto a school, he moves to other areas 

 because circling by boat is believed to force the fish 

 down, rendering them difficult to catch for some time. 

 In future surveys, an algorithm incorporating maxi- 

 mum swimming speeds and surfacing behavior could 

 be used to limit redundant counts. Minimum counts 

 based on daily rather than pooled totals would also re- 

 duce counting problems resulting from residence time 

 and sequential movement offish through the study area. 



Conclusions 



In this collaborative study neither we nor the ECTA 

 believe that expended effort was sufficient to derive, 

 on any given day, a minimum count of giant and 

 large-medium bluefin tuna in New England waters. 

 To do so would require not only perfect environmen- 

 tal conditions (noted by pilots and fishers as a "show" 

 day) but also the complete cooperation and coordi- 

 nation of efforts by participating pilots. Given that the 

 feasibility survey started relatively late in the fishing 

 season, the latter was a difficult goal to achieve. 



A fundamental limitation of aerial assessment is 

 that only fish at or near the surface are accounted 

 for, providing only a minimum estimate of school size. 

 Surveyed schools subsequently captured by seiners 

 might help define relations between aerial counts and 

 total biomass, but this relation was established for 

 only one set in the 1993 season. In this case ( 27 Aug), 

 we counted 32 fish at the surface of a tightly domed 

 school that yielded 125 large giants once captured. 



Factors that govern schooling behavior and aggre- 

 gation dynamics are poorly documented in bluefin 

 tuna and the tunas in general (Mather, 1962; Clark 

 and Mangel, 1979; Partridge et al., 1983). Mather 

 (1962) described bluefin tuna behavior patterns of 

 "pushing, milling, and smashing" in New England 

 waters. We have noted spatial configurations (domes, 

 cartwheels, surface sheets) and soldier groups (Par- 

 tridge et al., 1983) depicted in photographed schools. 

 Understanding the interplay of ecological and envi- 

 ronmental factors that govern aggregation of giant 

 bluefin tuna would help to define biases in direct 

 assessment efforts. For example, giant bluefin tuna 



