SMITH: SIZE CHANGES OF THREE DOLPHIN POPULATIONS 



full use was not evident even by 1964. Comparing 

 kill rates with and without "backdown" is compli- 

 cated however, because the effectiveness of the 

 release procedure has increased over time. 



No information is available on kill rates from 

 non-U. S. vessels during 1959-72, but the non- 

 U.S. fleet was small. There is little reason to sus- 

 pect that these kill rates were different, because 

 fishermen of both fleets were still learning how 

 to use purse seine gear for catching tuna in asso- 

 ciation with dolphins and how to release the 

 caught dolphins. 



The available kill rate data for this period 

 were stratified, for use in Equation (6), by 

 amount of tuna caught, size of the vessel, and fre- 

 quency of use of the "backdown" procedure. The 

 data were pooled across the years 1964-72 and 

 extrapolated back to the years 1959-63 when no 

 kill rate data were collected. These stratified kill 

 rates were multiplied by the number of sets 

 made on dolphins in each stratum to estimate the 

 total number of dolphins, of all populations, killed 

 directly in this fishery. 



Estimating proportions of the total kill of dol- 

 phins from each population for this period is dif- 

 ficult because the yellowfin tuna purse seining 

 was expanding westward and because data on 

 the species of dolphins observed killed are avail- 

 able only for 1971 and 1972. Prior to 1969 this 

 fishery operated shoreward of the range of the 

 whitebelly spinner dolphin, primarily within the 

 range of the eastern spinner and offshore spotted 

 dolphins. The total kill estimates are prorated to 

 population for the years 1959-72, based on ob- 

 served proportions in the 1971-72 data of 70, 23, 

 and 3% for offshore spotted, eastern spinner, and 

 whitebelly spinner dolphins, respectively. The 

 other 4% consisted of several species, primarily 

 common dolphins, Delphinus delphis, which are 

 not considered in this study. Although the tuna 

 purse seine fishery was expanding seaward 

 throughout the 1960's toward the range of the 

 whitebelly spinner dolphin, a major seaward 

 shift occurred in 1969. Lacking detailed data, I 

 assume this year to be the first significant in- 

 volvement of the whitebelly population. 



Some additional data on the species of dolphins 

 involved in each set has recently become avail- 

 able from the IATTC, suggesting a declining 

 proportion of sets involving spinner dolphins and 

 an increasing proportion involving spotted dol- 

 phins throughout the 1960's. Preliminary exami- 

 nation of these data indicates that the overall 

 proportions of sets involving each species are not 



greatly different from the 1971-72 observer data. 

 Direct use of these new data will involve making 

 a number of assumptions about species-specific 

 kill rates. 



Using the above proportions based on the 1971- 

 72 data and increasing the estimates of total 

 number killed by 4.8% to account for those dol- 

 phins possibly dying of injuries, I estimated the 

 total numbers of dolphins killed, by population 

 (Table 4). These are revisions of estimates used 

 by the 1979 workshop (footnote 3). 



Table 4. — Estimates of numbers (in 

 thousands) of dolphins killed by all fleets 

 in the eastern tropical Pacific, 1959-78, 

 for three populations of dolphins (Smith 

 text footnote 3). 



For the Period 1973-80 



Substantially more data exist on kill rates for 

 the period 1973-78 than for the period 1959-72. 

 The 1973-78 data are more reliable because they 

 were collected by NMFS employees trained spe- 

 cifically for obtaining kill information. Starting 

 in 1974 fishing trips were randomly selected for 

 observation to obtain a representative sample. 

 Greater cooperation by the fishing fleet resulted 

 in an increasing proportion of selected trips 

 actually observed from 1974 to 1976. However, 

 it was not until 1976 that fishing trips begun 

 after July were sampled. In the early 1970's fish- 

 ing tended to occur farther offshore later in the 

 year; because kill rates are generally higher in 

 the offshore areas, the failure to collect data from 

 late-season trips probably resulted in an under- 

 estimate of actual dolphin kill rates in those 

 years. This problem is partially accounted for by 

 stratifying the data by area. The species composi- 



