Reilly and Fiedler: Interannual variability of dolphin habitats 



435 



pattern of seasonal change in the ETP, associated 

 with the north-south movement of the Intertropical 

 Convergence Zone between the trade winds (Fiedler, 

 1992). Statistical tests supported the habitat hypoth- 

 eses of Au and Perryman ( 1985 ) for spotted, spinner, 



"J 



Spotted & Spinner Do 



J I I I I I L__ 



T 1 1 n 1 1 T 



^ o o o 



Striped Dolphins 



J I I I I I I L 



160 155 150 145 140 135 1J0 125 120 115 110 105 100 95 90 



Longitude 



Figure 1 



Dolphin distributions in the northern summer in the eastern tropi- 

 cal Pacific, from Reilly (1990). Contours are for encounter rates 

 per 185 km searched. Dots represent centers of 2° squares in which 

 there was at least 185 km search effort. 



and common dolphins, but not for striped dolphins. 

 During the summer, striped dolphins occupied habi- 

 tat that was geographically complementary with 

 habitats of both common and of spotted/spinner dol- 

 phins (Fig. 1). Striped dolphin habitat was indistin- 

 guishable statistically from either the upwelling- 

 modified or tropical habitats with the vari- 

 ables used, indicating that other factors 

 or processes act to separate these dolphins 

 from the others. 



Interannual variation in the region's 

 oceanography is dominated by the quasi- 

 periodic El Nino-Southern Oscillation 

 (ENSO; Enfield, 1989). Interannual varia- 

 tion exceeds seasonal variation in much 

 of the ETP (Fiedler, 1992). ENSO variabil- 

 ity affects all of the patterns and processes 

 so far identified with cetacean habitats in 

 the ETP. From this we hypothesized that 

 dolphin distributions might change 

 interannually in response to changes in 

 the distribution of their habitats. If so, this 

 could have important implications for the 

 monitoring of trends in animal abundance 

 now in progress from research vessel sur- 

 veys (Gerrodette and Wade, 1991) and 

 tuna vessel observer data ( Anganuzzi and 

 Buckland, 1989). 



The primary objective of this study was 

 to examine interannual variability in dol- 

 phin habitats, defined by multivariate 

 techniques, for the years 1986-90. We also 

 suggest methods for the use of habitat in- 

 formation in the monitoring of trends in 

 cetacean abundance. That is, given a quan- 

 titative definition of habitat and a record 

 of interannual changes in habitat distri- 

 bution, can we help reduce variance or 

 bias, or both, in the estimation of abun- 

 dance or tests for trends? 



Study area 



The basic physical features of the upper 

 ocean in the eastern tropical Pacific have 

 been described by Wyrtki ( 1966, 1967), and 

 Tsuchiya (1974). Fiedler (1992) updated 

 this description and summarized seasonal 

 and interannual variability. Major surface 

 water masses and currents are depicted 

 in Figure 2. Warm, low-salinity tropical 

 surface water is found in the center of the 

 ETP. Cooler, higher-salinity equatorial 

 surface water is found south of about lat. 



85 80 75 



