1976 as the maximum allowable kill by the U.S. 

 tuna fishery. Methods for monitoring the mortal- 

 ity levels, and projecting when the quota would 

 be reached during the year, were required. A 

 yellowfin tuna quota(175,000-195,000 short tons) 

 managed by the Inter-American Tropical Tuna 

 Commission (IATTC), around which the tuna 

 fishermen planned their fishing operations, was 

 also in effect and had to be incorporated into the 

 procedure. In this paper, we describe statistical 

 methods which have been used to estimate the 

 annual incidental dolphin mortality for the U.S. 

 fleet during the year and at the end of the year 

 since 1976. This estimative procedure has been 

 used also for foreign fleets by IATTC since 1979 

 (Allen and Goldsmith 1981). 



Methods 



The data sources used to monitor and estimate 

 incidental dolphin mortality were the scientific 

 observer program of the NMFS and the logbook 

 records of the IATTC. 



The NMFS observer program provides data 

 on discarded dolphins. Trained technicians were 

 placed aboard a random sample of U.S. tuna ves- 

 sels to collect data of various types, including 

 number of dolphins killed in a set, amount of 

 tuna caught, species of tuna, fishing location, 

 vessel capacity, and duration of trip. 



The IATTC maintains a logbook system where- 

 by it collects data on type of set, fishing locations, 

 tonnage of catch, species of tuna, vessel carrying 

 capacity, and other information that are recorded 

 in logbooks by fishermen. 



Three mortality rates were used to estimate 

 total dolphin mortality. They were obtained by 

 dividing the total observed kill of dolphins by the 

 total observed number of dolphin sets (kill-per- 

 set), by the total observed number of days-at-sea 

 (kill-per-day), and by the observed total catch of 

 yellowfin tuna associated with dolphin (kill-per- 

 ton). 



Estimation Procedures 



Three estimation procedures were used in this 

 study. The first procedure was based on kill-per- 

 day statistics to monitor the dolphin mortality 

 during the year; the second was based on kill- 

 per-ton combined with kill-per-day to project the 

 closure date; and the third was based on kill-per- 

 set to estimate the total mortality at the end of the 

 year. 



The kill-per-day and the kill-per-set methods 

 were based on stratified ratio estimators. Trips 

 from which the dolphin set data were taken, 

 were stratified according to fishing locality, ves- 

 sel carrying capacity, yellowfin tuna catch, gear 

 type, and fishing time. The fishing locality and 

 time were directly related to the I ATTC's yellow- 

 fin tuna regulatory system, which includes 1) an 

 annual quota on yellowfin tuna catch within the 

 Commission's yellowfin regulatory area(CYRA) 

 (Fig. 1), 2) season closure to enforce the quota 

 on yellowfin tuna catch, and 3) a "last trip" allo- 

 cated at season closure to each vessel that fished 

 during the open period (Table 1). 



Kill-Per-Day Method 



The kill-per-day method requires all trips to be 

 stratified according to gear type, vessel carrying 

 capacity, and fishing time; this method was de- 

 veloped to monitor kill during the season. The 

 year was divided into three periods, each desig- 

 nating a trip type. Trip-type 1 included all vessel 



Figure 1.— The major part of the Inter-American Tropical 

 Tuna Commission Yellowfin Regulatory Area (Courtesy of 

 IATTC). 



397 



