HOLT ET AL.: MONITORING DOLPHIN ABUNDANCE 



so H 70 tt 



IX M 110 H 



LONOITUDE 



were substituted for the respective stratified esti- 

 mates. 



Specific formulae to estimate variables and as- 

 sociated theoretical variances in Equations (2) 

 through (4) are from Burnham et al. (1980), Holt 

 (1985,8 in press) and Barlow and Holt (1986). 

 Variances for estimates of school sizes and school 

 densities were calculated using jackknife tech- 

 niques (Miller 1974). 



Since serial correlation among sampling units 



^Holt, R. S. 1985. Estimates of abundance of dolphin 

 stoclts taken incidentally in the eastern tropical Pacific yel- 

 lowfin tuna fishery. Southwest Fish. Cent. Adm. Rep. No. LJ- 

 85-20, 32 p. SWFC La Jolla Laboratory, National Marine 

 Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. 



(days of effort) will yield biased estimates of 

 standard errors using the jackknife method, we 

 analyzed serial correlation of dolphin school de- 

 tection rates among various combinations of suc- 

 cessive days of effort. Analyses indicated that cor- 

 relation was significant among successive single 

 days but was not significant for periods of 2 or 

 more days. Therefore, the data were grouped by 

 2-d increments for the jackknife analyses. 



Estimates of spotted dolphin population abun- 

 dance and values used in Equations (2) and (3) to 

 calculate the estimates are presented in Table 1. 

 CV (N)s were smaller for estimates calculated 

 using method B than for estimates using method 

 A. 



439 



