Julian and Beeson: Estimates of marine mammal, turtle, and seabird mortality 
275 
Table 3 
Observed (obs) and estimated (est) cetacean, pinniped, turtle, and seabird entanglement, stratified by year, in the California swordfish 
and shark drift gillnet fishery during the NMFS Observer Program, July 1990-December 1995. Estimates of entanglement are 
reported to the nearest individual. Estimated coefficients of variation (CV) are included in parentheses; ( — ) indicates CV was 
undefined. Effort and estimates for 1990 pertain to the third and fourth quarters only. 
Year 1990 1991 1992 1993 1994 1995 
obs 
est 
CV 
obs 
est 
CV 
obs 
est 
CV 
obs 
est 
CV 
obs 
est 
CV 
obs 
est 
CV 
Common dolphin 
(unknown stock) 
4 
92 
(0.79) 
7 
71 
(0.70) 
6 
44 
(0.36) 
4 
30 
(0.57) 
1 
6 
(0.91) 
0 
0 
(— ) 
Cuvier’s beaked whale 
0 
0 
(— ) 
0 
0 
(— ) 
6 
44 
(0.36) 
3 
22 
(0.53) 
6 
34 
(0.36) 
6 
39 
(0.36) 
Sperm whale 
0 
0 
(— ) 
0 
0 
(— ) 
3 
22 
(0.94) 
3 
22 
(0.69) 
0 
0 
(— ) 
0 
0 
(— ) 
Humpback whale 
0 
0 
(— ) 
0 
0 
(— ) 
0 
0 
(— ) 
0 
0 
(— ) 
1 
6 
(0.91) 
0 
0 
(— ) 
California sea lion 
2 
46 
(0.99) 
4 
41 
(0.58) 
9 
66 
(0.34) 12 
90 
(0.39) 
5 
28 
(0.40) 
5 
32 
(0.40) 
Unidentified sea lion 
2 
46 
(0.97) 
1 
10 
(0.95) 
0 
0 
(— ) 
0 
0 
(— ) 
0 
0 
(—) 
0 
0 
(— ) 
Loggerhead turtle 
0 
0 
(— ) 
0 
0 
(— ) 
2 
15 
(0.66) 
5 
37 
(0.49) 
0 
0 
(— ) 
0 
0 
(— ) 
Leatherback turtle 
1 
23 
(0.97) 
1 
10 
(0.94) 
4 
29 
(0.46) 
3 
22 
(0.53) 
1 
6 
(0.91) 
5 
32 
(0.47) 
Unidentified turtle 
0 
0 
(— ) 
0 
0 
(— ) 
0 
0 
(— ) 
3 
22 
(0.93) 
0 
0 
(— ) 
0 
0 
(— ) 
Seabirds (all unidentified) 
1 
23 
(0.98) 
0 
0 
(— ) 
1 
7 
(0.93) 
0 
0 
(— ) 
1 
6 
(0.90) 
0 
0 
(— ) 
3 Underestimation of effort from landing receipts. 
This type of error may occur when a landing re- 
ceipt, in absence of additional information, is as- 
sumed to represent one effort-day (default- value). 
A reliable determination of the magnitude of bias 
these errors cause in estimates of mortality rates has 
not been made; however, their characteristics indi- 
cate that estimates of mortality may be lower than 
actual values. 
Once quarterly effort data were collected, computer 
programs developed by CDFG were used to assign 
target species to landing receipts on the basis of in- 
formation provided by logbooks and observations 
(Beeson and Hanan 9 ). Landing data were then con- 
firmed or modified on the basis of logbook and ob- 
server data for the same target species, date, and 
vessel number. After all three data sources were com- 
pared, a day of effort was tallied for each record with 
a logbook entry or observer record. Landing receipts 
without corresponding logbook or observer entries 
three days before and after the receipt date were 
assumed to represent one day of effort. The num- 
bers of days fished in each CDFG block were then 
tallied and the resultant data represented estimated 
total effort. Total effort was estimated quarterly and 
yearly. Delayed submission of data to CDFG was the 
primary reason that estimates of yearly effort differed 
from the sum of the quarterly estimates of effort. 
9 Beeson, M., and D. Hanan. 1996. Manuscript submitted to 
California Fish and Game. [Available from the authors at Cali- 
fornia Department of Fish and Game, 330 Golden Shore, Suite 
50, Long Beach, CA 90802.] 
Mortality estimation 
Sampling design The NMFS observer program 
began in July 1990. Initially, the plan was to sample 
every fifth trip made by a drift gillnet vessel accord- 
ing to an assignment schedule. The order in which 
vessels with MMPA exemption certificates were 
scheduled to be observed was randomly selected. It 
became evident during the 1990 season that this 
scheme would not work because of logistical difficul- 
ties in adhering to the sampling plan. 
Beginning in January 1991, gillnet vessel trips 
were selected according to the targeted coverage rate 
(20%), the availability of observer personnel, call-ins 
(fishermen called in prior to departure), and the abil- 
ity to notify fishermen of their obligation to carry an 
observer. (Occasionally fishermen did not call the 
Observer Program administrator for possible ob- 
server assignment.) In addition, the NMFS Fisher- 
ies Observer Branch began monitoring vessel activ- 
ity (arrivals and departures) to estimate observer 
coverage for placement purposes. If estimated ob- 
server coverage dropped below 20% for a vessel, the 
owner was notified of the obligation to carry an ob- 
server. In the setnet fishery, most fishermen were 
notified after their nets were set whether they would 
be required to carry an observer. As the program 
evolved, setnet fishermen began to expect an observer 
about every fifth trip. 
Mortality estimation in the drift gillnet fishery In 
the swordfish and shark drift gillnet fishery, vessels 
made trips lasting from one to about 15 days so that 
