Preliminary Report 
Pelagic Non-Grid Observations 
Southern Island Cruises 17 & 18 
January-April 19&7 
by 
Richard D. Chandler 
INTRODUCTION AND PREFACING REMARKS 
A. General 
Southern Island Cruises (SIC) 17 and 18 are the last in a current 
series of ecological investigations of the Equatorial Central Pacific 
waters and islands. This report summarizes four months of shipboard ob¬ 
servations of the birds encountered in a roughly triangular area of ap¬ 
proximately one million square miles delimited by the three islands, 
Oahu (Hawaii), Starbuck (south end of the Line Islands), and Viti Levu 
(Fiji Islands). 
The two cruises are ideally treated as a unit for several reasons. 
Primarily, the combined coverage of the two trips overlaps all major areas 
previously investigated on other SIC trips. Also, both trips were made 
aboard nearly identical ships and nearly the same POBSP personnel were on 
both cruises. Standardized techniques and methods progressively learned 
during three earlier cruises to this area in the preceding six months give 
the combined data from these two final cruises a higher degree of com¬ 
parative constancy. 
B. Data Collection 
Four questions can be asked about faunal distribution in general: 
What is it?, Where is it?, When is it there?, Why is it there?. In the 
at-sea bird surveys the first three questions are answered straight away 
whenever a positive identification of a bird is recorded along with its 
coordinates in time and space. Stating, "That Sooty Tern is right here, 
right now” operationally defines its distribution at that moment. However 
the statement applies only to the individual bird. Ii it is the distribu¬ 
tion of the species that is of interest then another operative deiinition 
must be made. An "instantaneous spatial distribution" could be defined as 
"the set of all spatial coordinates corresponding to each individual of 
the species at a given time" -- a map sprinkled with dots. For some pur¬ 
poses this is useful information, but having no time variation there is no 
dynamic information to be gathered. A series of such "map sprinklings" 
taken at specified time intervals, however, yields much in the way of 
dynamic information. In the limit, a "species" distribution can be opera¬ 
tionally defined as "the sequential set of all possible instantaneous 
spatial distribution/* Since species is not defined on a temporal scale 
this already absurd definition is not useful because of a definition con¬ 
flict. A useful definition of "distribution" results from the value 
placed on the degree of confidence desired in the final answer; that is, 
how close to the above limit one needs to go before the information yiej_a 
is not worth the expenditure. 
