Edwards and Perkins: Estimated tuna discard from dolphin, school, and log sets 
221 
tained losses of even 30% annually might pose long- 
term problems for the fishery. 
Caveats 
Our results are based on a number of assumptions, 
including among others 1) similarity between effort 
patterns of the U.S. and international fleets in the 
ETP and 2) constancy in discard per set within each 
set type, regardless of changes in ecological or eco- 
nomic forces (e.g. oceanographic conditions, fish 
prices, or both), proportions of effort devoted to each 
set type, or times and areas fished. In addition, our 
study does not address possible limitations on the 
availability of nondolphin sets, e.g. limitations on the 
number of available schools or logs, possible changes 
in fleet sizes with the elimination of dolphin sets (e.g. 
seiners leaving the ETP), or the possibility of greater 
emphasis on log fishing through increased use of 
FAD’s (fish aggregating devices). Our analysis also 
assumes that the catch per vessel, the number of 
vessels in the fleet, and the geographic distribution 
of set types would remain constant regardless of the 
amount of effort expended within each set type. 
One or more of these assumptions may be invalid 
(e.g. Punsly et al., 1994), but at present the effects of 
violating them cannot be assessed quantitatively. For 
example, relatively greater effort in area 1 would 
increase tuna discard more than we have estimated 
in the present study, but increased competition for 
logs might decrease overall fishing success. Con- 
versely, increased availability of “logs” (where the 
population of logs has been increased by deployment 
of FAD’s) could mitigate this competition and lead to 
an increase in fishing success. Conversely yet again, 
increased availability of logs might “dilute” the avail- 
able tuna resource per log, leading to decreased fish- 
ing success. The actual system response cannot be 
ascertained at this time. 
In addition, the fine-scale size structure of the dis- 
cards, particularly from log sets, is unknown. The 
estimated number of individuals potentially lost per 
short ton varies by a factor of three; at the lower es- 
timate, only about 8% of recruits would be lost to 
discard; at the higher estimate, closer to 25% would 
be lost. More estimates await subsampling studies 
15 continued 
the predicted redistribution of log and school effort to those ar- 
eas might not occur. Instead, the fishery would likely contract 
to more nearshore areas, similar to the situation during the 
1960’s. Yellowfin tuna resident in the offshore zones would es- 
cape all exploitation with the resulting appearance of a much 
smaller stock — on the order of another 25% reduction in recruit- 
ment (Deriso, R. 1995. IATTC, c/o Scripps Institute of Ocean- 
ography, 8604 La Jolla Shores Drive, La Jolla, CA, 92038. Per- 
sonal commun.). 
not yet completed. However, even a moderate esti- 
mate of 15% loss per year is difficult to dismiss as 
unimportant, particularly in combination with the 
significant increase in landed recruits that will re- 
sult from increased log fishing. 
More detailed data on size structure of the by catch 
would also make it possible to address the issue of 
changes in the discard of medium to large yellowfin 
tuna that might occur in response to changes in ef- 
fort distribution. Large and medium yellowfin can 
only become discard in the event that they grow to 
medium or large size; increases in log fishing will 
decrease survival of young yellowfin tuna and de- 
crease the probability that these fish will show up 
later either as direct catch or as discards. Such de- 
tailed size-structure data would also make it possible 
to translate all discard data to a standardized age, 
making it easier to compare discards between size 
classes (e.g. a single 2-year-old yellowfin tuna might 
be found to be equivalent to 100 or more 6-month- 
old fish if mortality remained the assumed constant 
average over time). These more detailed studies 
should be possible in future given the intensive dis- 
card sampling program initiated recently by the 
IATTC (Deriso). 16 
Conclusions 
The work presented here provides only a general 
overview of the apparent situation based on data 
available to date, but the implications of our simple 
study are serious and merit further consideration. 
Our calculations, although only approximate, imply 
strongly that a change from dolphin to school or (es- 
pecially) log fishing (or both), although it might alle- 
viate the problem of dolphin mortality, could have 
significantly detrimental effects on other components 
of the eastern tropical Pacific Ocean ecosystem, not 
the least of which is the commercial resource upon 
which the fishery rests. 
Acknowledgments 
We thank Al Jackson, for his help and advice about 
data collection methods and data form interpreta- 
tion, and the Inter- American Tropical Tuna Commis- 
sion, for generously making available their summary 
data on tuna discards from the ETP tuna purse-seine 
fleet. We also thank Richard Deriso, Martin Hall, and 
16 Deriso, R. 1996. IATTC, c/o Scripps Institute of Oceanogra- 
phy, 8604 La Jolla Shores Road, La Jolla, C A 92038. Personal 
commun. 
