FISHERY BULLETIN: VOL. 81. NO. 1 



tion of the kill was also recorded, allowing direct 

 estimates of total kill of dolphins, from Equation 

 (6), for each population. 



The number of dolphins killed per set from 

 1973 to 1976 for successful and unsuccessful sets 

 was about 18 and 3, respectively, a decrease from 

 the 1964-72 levels of 56 and 8. The number killed 

 in successful and unsuccessful sets in 1977 and 

 1978 was again lower, about 4 and 2, respectively 

 (Table 2). These decreases occurred as U.S. regu- 

 lations were developed and eventually imple- 

 mented, and as methods for more effective use of 

 backdown and other dolphin-release procedures 

 were developed and used. The decreases in kill 

 rates were apparently due, at least in part, to 

 wider adoption of procedures for dolphin release. 



The non-U. S. tuna purse seine fleet increased 

 markedly during this period. First observations 

 of the kill rate for this fleet were in 1979, which 

 showed that the rate was very similar to that of 

 the U.S. fleet (Allen and Goldsmith 1981). Given 

 this similarity in 1979, it is reasonable to assume 

 that during the earlier part of the 1970's the non- 

 U.S. kill rate declined, as did the U.S. kill rate 

 (Table 2), as dolphin-release technology devel- 

 oped by the U.S. fleet became known. If such a 

 decline in the non-U. S. kill rate occurred, how- 

 ever, it would probably have been somewhat 

 slower than that for the U.S. fleet, because of 

 lack of legal pressure to reduce the incidental kill 

 and time lags in technology transfer. Following 

 the procedure developed by the 1979 workshop 

 (footnote 3), I estimated the non-U. S. kill by 

 assuming 1) the same kill rate in 1971-72 for the 

 non-U. S. fleet as that observed aboard U.S. ves- 

 sels in those years; 2) the same kill rate in 1973 

 for the non-U. S. fleet as that of the U.S. fleet in 

 1975; and 3) a linear convergence of the two rates 

 toward the 1979 U.S. rate. Estimates of numbers 

 of dolphins killed by non-U. S. vessels obtained 

 under these assumptions are used here. How- 

 ever, additional study is needed, especially since 

 the recorded kill rate for the non-U. S. fleet in 

 1980 was somewhat higher than that for the U.S. 

 fleet (Allen and Goldsmith 1982). 



These kill rates, stratified by vessel size, 

 amount of tuna caught, and area fished, are used 

 in Equation (6), along with the estimated num- 

 ber of sets on dolphins, to estimate total direct 

 kill by population for each year. These estimates 

 are then increased by 4.8% to account for dol- 

 phins assumed to die of their injuries (Table 4). 

 The results in Table 4 are slight revisions of the 

 estimates used by the 1979 workshop (footnote 3). 



NET RECRUITMENT RATE 

 ESTIMATES 



Maximum net recruitment rate (i?,„) is re- 

 quired to estimate historical abundance. This is 

 calculated as the difference between gross pro- 

 duction of calves and the natural mortality rate, 

 assuming that natural mortality does not change, 

 when a population is reduced substantially be- 

 low its equilibrium level. 



Gross Reproductive Rates 



Gross recruitment rates can be estimated as 

 the product of the female fraction of the popula- 

 tion, the mature female fraction, and the annual 

 pregnancy rates. Estimates of these parameters 

 are given in Table 5, based on samples of dol- 

 phins collected by scientific observers aboard 

 tuna vessels from 1973 to 1978. Two methods 

 were used to estimate the annual pregnancy rate: 

 The first method (I) is the observed proportion of 

 pregnant females in the population divided by 

 the gestation period; the second method (II) is 

 similar, but uses additional information on fre- 

 quency of nursing calves in the samples from 

 each net set (Perrin et al. 1977a, b, c). 



There are known sampling biases in these data 

 for spotted dolphin because of the fishing pro- 

 cess, partly accounted for by using data for spot- 

 ted dolphin recruitment rates from only those 

 sets where more than 40 dolphins were killed. In 

 addition, the observed fraction of the mature, 

 pregnant female dolphins has varied among 

 years, with a general decline in offshore spotted 

 dolphin and a large degree of variability in east- 

 ern spinner dolphin. 



Age-specific effects are not accounted for in 

 the analyses so far, however, particularly the 



Table 5. — Proportion of sampled dolphins (female and ma- 

 ture) of three populations and estimates of annual pregnancy 

 rate (P) and gross reproductive rate (G), using two methods. 1 

 See text for details. 



Annual production 



'Henderson, J. R..W. F. Perrin, and R. B. Miller. 1980. Ratesof gross 

 annual production in dolphin populations (Stenella spp and Delphinus 

 delphis) in the eastern tropical Pacific, 1973-1978 Southwest Fish. 

 Cent. La Jolla Lab., Natl. Mar Fish. Serv , NOAA. Admin Rep. LJ-80-02. 

 51 p. 



8 



