616 



Fishery Bulletin 91(4), 1993 



although a Kolmogorov-Smirnov test to compare the 

 age-frequency distributions of sexually immature and 

 sexually mature animals from the two studies con- 

 cluded that the samples were likely drawn from the 

 same population. The underrepresentation of the 

 immature age classes may be a result of segregation 

 in the population by age and sex, fairly typical in large 

 mammal populations, and may affect the estimation 

 of ASM. We investigated the effect of this un- 

 derrepresentation by assuming a stable age distribu- 

 tion and by apportioning the animals in the indeter- 

 minate age classes by sexual maturity under different 

 assumptions of sexual maturity for animals in those 

 age classes, and then calculated ASM. On the basis of 

 segregation by age and sex observed in other species 

 of large mammals, we predicted that the most reason- 

 able scenario for reconstructing the age distribution is 

 that the "missing" animals would be sexually imma- 

 ture. If this indeed were the case, we have underesti- 

 mated ASM. Under this and other scenarios we inves- 

 tigated, estimates of ASM were consistently over- or 

 under-estimated for both stocks. Therefore, possible 

 sampling biases were unlikely to be responsible for 

 the observed differences in ASM between the northern 

 offshore and southern offshore stocks of spotted 

 dolphin. 



The hypotheses we tested required that several im- 

 plicit assumptions be made, including: 1) constant data 

 collection biases, 2) minimal interchange of animals 

 between areas, 3) consistent estimation of specimen 

 age, 4) linear compensatory response with change in 

 population abundance, 5) constant carrying capacity 

 (K), and 6) equivalent initial life history parameters 

 for the two stocks. The first three of the assumptions 

 are believed to hold reasonably well. Barlow (1985) 

 tested several of the same life history parameters we 

 examined and found them to be relatively insensitive 

 to a number of potential data collection biases; no ma- 

 jor changes in the collection of life history data were 

 made after 1974, thus supporting constant data collec- 

 tion bias as a reasonable assumption. Furthermore, 

 the fishery operates in the same areas from year-to- 

 year at approximately the same time of year. In fact, 

 60% of our sample from the northern offshore stock 

 was collected between July and September, and a fur- 

 ther 38% was collected between April and June; in the 

 south, 57% of the sample was collected between Janu- 

 ary and March, and 27% was collected between Octo- 

 ber and December. This pattern of sample collection 

 was consistent from year-to-year, and therefore would 

 not affect the pooled or annual estimates of ASM or 

 reproductive parameters. Geographic variability has 

 been noted in earlier studies of spotted dolphin life 

 history data (Hohn and Hammond, 1985; Barlow, 1985); 



thus selecting specimens from discrete areas as we did 

 reduces the potential for geographic variability to ob- 

 scure population compensatory responses. Similarly, 

 estimating the age of all specimens at one time and in 

 the blind ensured that age estimates were made as 

 consistently as possible. Information about the valid- 

 ity of assumptions 4) and 5) is not currently available, 

 but violations of the assumptions may provide expla- 

 nations for our results. There is evidence for differ- 

 ences in the morphological and life history character- 

 istics between the northern offshore and southern 

 offshore stocks of spotted dolphin (see Perrin et al., 

 1976, 1979b, 199F; Barlow, 1985; Hohn and Hammond, 

 1985; Myrick et al, 1986; Bright and Chivers 4 ) that 

 are likely correlated with the different oceanographic 

 environments of the areas inhabited by these stocks 

 (Au and Perryman, 1985; Reilly, 1990). 



We did not find conclusive evidence for compensa- 

 tory responses in these stocks of spotted dolphin as 

 only one reproductive parameter for each stock showed 

 a statistically significant trend. However, observed 

 trends in both parameters suggest the populations are 

 below K and declining. Our comparison of ASM's sug- 

 gest further biological differences between the north- 

 ern offshore and southern offshore stocks or popula- 

 tions. Currently, the order of compensatory responses 

 and their dynamics are not known. In order to be use- 

 ful as a biological index, a measure of the status of the 

 population must be known and the dynamics of the 

 parameter over a wide range of population densities 

 quantified (Fowler and Siniff, 1992). 



Acknowledgments 



We thank the NMFS and IATTC biological technicians 

 who collected the life history data while aboard the 

 tuna purse-seiners, often under difficult working con- 

 ditions, and the Southwest Fisheries Science Center 

 personnel that coded and edited the data for use. 

 Priscilla Akin and Andrea Bright prepared the teeth 

 for estimating the age of the specimens. We thank Jay 

 Barlow, Douglas DeMaster, Andrew Dizon, Christina 



'Perrin. W. F„ G. D. Schnell. D. J. Hough, J. W. Gilpatrick, and J. V. 

 Kashiwada. 1991. Re-examination of geographical cranial variation 

 in the pantropical spotted dolphin (Stenella attenuata) in the east- 

 ern Pacific. Dep. Cummer., NOAA, Natl. Mar. Fish. Serv., Southwest 

 Fish. Sci. Cent., P.O. Box 271, La Jolla, CA 92038. Admin. Rep. LJ- 

 91-39, 46 p. 



"Bright, A. M., and S. J. Chivers. 1991. Post-natal growth rates: a 

 comparison of northern and southern stocks of the offshore spotted 

 dolphin. Dep. Commer., NOAA, Natl. Mar. Fish. Serv., Southwest 

 Fish. Sci. Cent., P.O. Box 271, La Jolla, CA 92038. Admin. Rep. LJ- 

 91-30, 24 p. 



