4 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 123 



This work was supported by a National Science Foundation grant to 

 Dr. I. E. Wallen of the Smithsonian Institution in Washington as part 

 of the U.S. Program in Biology, International Indian Ocean Expedi- 

 tion. Drs. Philip S. Humphrey, Robert W. Storer, and George E. 

 Watson offered numerous helpful suggestions and encouragement 

 throughout the course of my Indian Ocean field activities and in the 

 preparation of this report. I am indebted also to all my companions 

 aboard the Anton Bruun for their assistance and independent con- 

 tributions and to Claudia L. McPeek, who handled the countless 

 administrative and other details connected with my participation 

 in this expedition. Roger S. Bailey provided stimulating discussion 

 of the problems of seabird distribution that are the concern of this 

 paper and made available many of his unpublished observations 

 and thoughts. John P. Hubbard and Guy Musser helped with the 

 illustrations. Finally, I am most grateful for the assistance rendered 

 by my wife Frances in the typing of the final draft of this manuscript. 



Seabird abundance. — Included in table 1 are the numbers of sea- 

 birds (grouped by order) observed during the standardized watch 

 periods and a calculated figure for the number of birds seen per hour 

 during each watch. The relationship of the latter figure to the actual 

 density of seabirds in a given area is biased by many variables in- 

 cluding the movement of the ship, movement of the birds, conditions 

 for observation, and perhaps time of day. The first of these may be 

 particularly important (compare Bailey, 1966, p. 261) since seabirds 

 rarely are distributed uniformly at sea but instead congregate at and 

 move with localized sources of food. Hence, any substantial move- 

 ment of the ship should increase the likelihood of encountering non- 

 randomly distributed seabirds. If this is true, the counts I made 

 during the morning watches on station are not directly comparable 

 to evening watches or others made when the ship was cruising at 

 normal speeds. An additional complication is the possibility that 

 the greater duration of the morning watches increased the chances of 

 seeing small groups of moving birds and partially compensated for 

 the lack of movement of the ship. Despite these and other problems 

 inherent in the quantification of at-sea observations, at least five 

 regions were found where seabirds were more common than in the 

 intervening areas. These were: (l) the western Arabian Sea and south 

 along the coast of Somalia; (2) the Seychelles Islands and surrounding 

 seas; (3) the waters near the Mascarene Islands; (4) subantarctic 

 waters below 33° S; (5) the seas east of the Maldive Islands and north 

 to the Indian coast. A sixth region near the equator also might be 

 distinguished though the increase there was less pronounced than 

 in the other areas. In these regions the number of seabirds seen per 

 hour was definitely greater than 2-3, usually 10-30, and on some occa- 



