HOLT: DENSITY OF DOLPHIN SCHOOLS 



better lateral viewing, but both provided unob- 

 structed forward and downward views. 



Two research vessels were used to collect the 

 shipboard data. The NOAA ship David Starr Jor- 

 dan was used during all years and the NOAA ship 

 Toivnsend Cromwell joined it in 1977, 1979, and 

 1980. Both vessels were similar in length and 

 cruising ability. Binoculars used to locate ani- 

 mals were mounted approximately 10.7 m above 

 the sea on the Jordan but were only 6.1 m above 

 the sea on the Cromwell. In addition, observers 

 aboard the Jordan used 20 x binoculars during 

 the 1977 surveys and 25 x glasses on the rest of 

 the surveys; observers aboard the Cromwell used 

 only 20 X glasses during their surveys. Conse- 

 quently, viewing conditions were generally much 

 better on the Jordan. 



Study Area 



Survey efforts traversed the combined range of 

 ETP dolphin stocks defined by Au et al. (1979)*^. 

 The range was partitioned into "inshore" and 

 "offshore" areas (Fig. 1). Airplanes were used to 

 survey the inshore area, and ship surveys were 

 conducted in both areas during each year, except 

 during 1977 when ships surveyed only the off- 

 shore area. 



Data Collection 



Aerial Data 



Data collecting procedures used during the 

 aerial surveys are described by SWFC (fn. 4), 

 Jackson (fn. 5), Holt and Powers (1982), and 

 Cologne and Holt (1984)'. As the airplanes tra- 

 versed predetermined tracklines (Fig. 1), the ob- 

 servers recorded schools on and to either side of 

 the lines. Observers searched through the bow 

 window and from windows located on either side 

 of the plane. The bow observer was responsible for 

 detecting schools on the trackline (a path under- 

 neath the plane 0.19 km wide). The searching 

 mode was halted if environmental or oceano- 

 graphic conditions restricted the observer's view 

 of the trackline or when the plane was diverted 

 from the trackline for closer examination of a 



6Au. D.. W. Ferryman, and W. Perrin. 1979. Dolphin dis- 

 tribution and the relationship to environmental features in the 

 eastern tropical Pacific. Southwest Fisheries Center Status of 

 Porpoise Stocks working paper SOPS/79'36, 59 p. 



^Cologne, J., and R. Holt. 1984. Observer effects in ship- 

 board sight surveys of dolphin abundance. Southwest Fish. 

 Cent. Adm. Rep. No. L-J-84-.30, 42 p. National Marine Fish- 

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



school. Additional schools detected during these 

 diversions were not included in the density analy- 

 sis. 



Sea conditions were measured on the Beaufort 

 scale (Bowditch 1966), which ranged from very 

 flat, glassy seas (Beaufort conditions) to rough 

 seas with numerous large, white-capped waves 

 (Beaufort 5 conditions). Sun location was de- 

 scribed by horizontal and vertical position rela- 

 tive to the bow observer (Holt 1983a). These were 

 recorded for each segment of effort. 



Biological and environmental data were 

 recorded at each sighting (Holt and Powers 1982). 

 Data included species identification, school size 

 estimates, sea state, sun position, and perpendic- 

 ular distance to the school from the trackline. 

 School size estimates consisted of an observer's 

 "best" estimate plus an estimate of the minimum 

 and maximum range. 



Ship Data 



Shipboard collection procedures are described 

 in the various cruise reports (unpublished docu- 

 ments available from the SWFC) and by Holt 

 (1983b). Procedures and data recorded on ship- 

 board surveys were similar to those for aerial sur- 

 veys. Two observers used binoculars located on 

 each side of the ship to search from directly ahead 

 to abeam of their respective sides of the ship. 

 Starting in 1979, sea state was recorded at the 

 beginning of each effort segment (leg). Sun posi- 

 tion was recorded during the 1982 and 1983 ship 

 surveys. 



The bearing (0) and radial distance (r) to a 

 school from the ship were recorded, and perpen- 

 dicular distance (y ) was then calculated as y = r 

 sin 9. In surveys conducted before 1980, observers 

 rounded estimates of sighting angles to multiples 

 of 5° or 10°, and radial distances to multiples of 

 185 m (0.1 nmi) within the first 1.85 km (1 nmi), 

 and to 0.93 km (0.5 nmi) multiples at larger dis- 

 tances (Fig. 2). During training, observers on the 

 1980 surveys were told of previous rounding inac- 

 curacies and instructed to make estimates as pre- 

 cise as possible. However, they were still unable 

 to make precise visual estimates of angles and 

 distances for schools recorded at great distances 

 from the ship (Fig. 2). During the 1982 and 1983 

 surveys, estimates of bearing were recorded using 

 a 360° graduated washer attached to the base of 

 the binoculars, and the radial distances were 

 measured using a graduated reticle enclosed in 

 the right eyepiece of the binoculars (Holt 1983b). 



421 



