220 
Fishery Bulletin 96(2), I 998 
yellowfin tuna can be applied to these set-specific 
tuna tonnages to produce estimates of 418, 1,540, and 
6,150 short tons of yellowfin tuna discarded annu- 
ally by the international fleet, under the assumption 
of observed U.S. effort patterns (Table 2). These esti- 
mates compare reasonably well with estimates of to- 
tal short tons of yellowfin tuna discarded by the in- 
ternational fleet during 1993 and 1994 of 449-917 
short tons from dolphin sets, 606-2,108 short tons 
from school sets, and 3,802-4,150 short tons from 
log sets (IATTC 14 ). 
Yellowfin tuna discards under redistributed pat- 
terns of effort can be predicted from tonnages pre- 
dicted under the different effort regimes. The “worst- 
case” discards will obviously occur when all dolphin 
sets are redirected to log sets. The total tonnage of 
tuna discard in this circumstance is the sum of dis- 
cards from dolphin sets (by definition 0, because we 
redistributed the dolphin sets effort to log sets), school 
sets (3,080 short tons, the same as under observed 
effort patterns because school effort is assumed to 
be unchanged), and log sets (103,000 short tons; es- 
timated as log discard from observed patterns, 28,000 
short tons, multiplied by 3.69, the increase factor for 
discard observed in U.S. data when redistributing all 
dolphin sets to log sets). The tonnage of yellowfin tuna 
in each of these set categories is 0 x 0.92 = 0 short tons 
for dolphin sets, 3080 x 0.5 = 1,540 short tons for school 
sets, and 103,000 x 0.22 = 22,700 short tons for log sets, 
with the “fraction yellowfin” from IATTC’s data base. 
Obviously, changing entirely from dolphin sets to 
an equal number of log sets dramatically increases 
predicted total bycatch of yellowfin tuna. A particu- 
larly unfortunate feature of this increase is the num- 
ber of small yellowfin tuna involved; the vast major- 
ity of log-set yellowfin tuna discards are new recruits 
to the fishery that have grown just large enough to 
be caught in the seine (Table 2). This increase in yel- 
lowfin tuna bycatch is unfortunate for two reasons: 
first, removing these small fish removes their poten- 
tial to grow larger and more valuable, thus reducing 
overall yield per recruit for the fishery (Punsly et 
al., 1994); second, these small fish are prereproduc- 
tive. No stock-recruitment relationship has yet been 
demonstrated for yellowfin tuna in the ETP, but if 
one exists, these increased losses to bycatch might 
become troublesome over long periods of time. 
The actual effect of these discards of small fish on 
future yield per recruit depends on the number dis- 
carded in comparison with the number recruited. 
Fortunately, the annual recruitment of yellowfin tuna 
14 IATTC Third Quarter Report. 1995. Table 10. Inter-Ameri- 
can Tropical Tuna Commission, c/o Scripps Institute of Ocean- 
ography, 8604 La Jolla Shores Drive, La Jolla, CA 92038. 
has been estimated by the IATTC. 5 The number of 
small fish discarded can be estimated from the weight 
discarded by assuming that all small discarded yel- 
lowfin are new recruits (Table 2). This is reasonable 
because the “small” size class of discards is defined 
in the IATTC bycatch data base as fish less than 2.5 
kg (Table 2), and Cole (1980) reported that yellowfin 
tuna recruit to the ETP tuna purse-seine fishery at 
sizes ranging from 0.7 to 2.0 kg. Thus a short ton of 
“small” discarded yellowfin tuna, at 907 kg per ton, 
should include between 454 and 1,296 individual fish. 
According to inferred effort patterns, estimated 
average annual discard of small yellowfin tuna by 
the international fleet was 4,950 short tons , repre- 
senting 2. 3-6. 4 million fish (Table 2). If all dolphin 
sets are assumed to be replaced by an equal number 
of log sets, then predicted discard of small yellowfin 
tuna increased to 17,562 short tons, representing 8.0— 
22.7 million fish (0 short tons from dolphin sets plus 
262 short tons from school sets plus 17,300 short tons 
from log sets). 
By comparison, the estimated annual recruitment 
of small yellowfin tuna to the ETP fishery for 1983- 
91 was 98 million fish (IATTC 5 ); thus observed ef- 
fort patterns may have been generating discards rep- 
resenting 2-7% of total recruitment each year. On 
the other hand, a redistribution of all dolphin-set ef- 
fort to log sets could increase discards to 8-24% of an- 
nual recruitment, depending on the actual sizes of the 
discarded fish ( smaller fish would lead to greater losses 
per ton of discard, although the increase in catchability 
with size in this range suggests that much of the dis- 
card might be closer to 2 kg than 0.7 kg). 
These estimated discards would occur in addition 
to an expected 25% reduction in yellowfin tuna catch 
biomass if all dolphin-fishing effort were redistrib- 
uted to log-fishing effort (Punsly et al., 1994). By 
implication, a change entirely to log fishing could 
reduce yellowfin tuna catch in the ETP by as much 
as 30-50%. A change to a combination of school and 
log fishing would be less detrimental, but the actual 
effect would depend on the ratio of school-set effort 
to log fishing effort. Historical patterns of purse-seine 
fishing in the ETP, in which 50% or more of effort 
was directed to dolphin sets, obviously minimized any 
removal of new recruits from the fishery. Replace- 
ment of the entire dolphin-set effort with log-set ef- 
fort would reverse that characteristic of the fishery, 
with potentially troublesome effects. Losses to “ob- 
served” recruitment as high as 50% may not be likely 
(Deriso 15 ), but it is not difficult to imagine that sus- 
15 It is also possible that a shift to log fishing could make some 
areas, particularly offshore areas, uneconomical to fish so that 
continued on next page 
