Total abundance for each species was esti- 

 mated on the basis of linear densities, taking 

 into account the effective range of observation 

 for each species and the area of the octants. 

 Figure 29 shows the densities of birds per 10 

 linear miles in each octant and the estimated 

 daily population in a given season within that 

 octant. The estimates of population in table 5 

 are for maximum numbers encountered in any 

 one month, whereas the estimates in figure 29 

 are averaged for a whole season. Albatross 

 density was based on cumulative totals through 

 the day rather than the highest number at one 

 time, the value used in the species account. 



Densities of total avifauna were highest in 

 spring, slightly lower in summer and fall, and 

 substantially lower in winter. The distribution 

 of high and low densities appears not to be at 

 random. Octant 6, which contained the land 

 masses, consistently contained the highest den- 

 sity of birds in all seasons but summer, when 

 it was second by a narrow margin. Octant 5, 

 the northwestern corner, varied considerably. 

 It ranked sixth in the fall after it ranked first 

 in the summer--which may well indicate that 

 Sooty Terns and Wedge-tailed Shearwaters 

 breeding in the main Hawaiian area do not dis- 

 perse in a northerly direction after their 

 breeding seasons. 



Octant 4, and to a lesser extent octant 8, at 

 the southern end of the study area, show con- 

 sistently high densities, especially in fall, 

 caused presumably by the attractiveness of the 

 area of relatively high productivity near lat. 

 10° N. where an underlying cold water dome 

 comes very close to the surface. Nonbreeding 

 populations of Sooty Terns, Wedge-tailed Shear- 

 waters, Juan Fernandez Petrels, and Black- 

 winged Petrels make up the bulk of the popula- 

 tions in octants 4 and 8. Octants 1, 2, and 3 

 were consistently low in avian density--espe- 

 cially octant 2, which ranked last or next to 

 last in all seasons. 



The following general patterns may be seen 

 in the analysis of density by octant: 



1. Birds that breed in the main or leeward 

 Hawaiian group in the spring and summer 

 are most abundant in octants 5 and 6 in 

 spring and summer. 



2. Birds that breed in the leeward Hawaiian 

 group in the winter attain highest den- 

 sities in octants 5 and 1 in winter and 

 spring. 



3. Birds that migrate rapidly in waves show 

 high densities in one or two octants and 

 low densities in all others. 



4. Birds that breed at a distance from the 

 study area and spend their nonbreeding 

 seasons in the study area attain highest 

 densities in octants 4 and 8. 



The estimates of total populations permit the 

 calculation of total bird-days, from which, with 

 certain assumptions about the food require- 

 ments of sea birds, can be derived the annual 

 consumption by sea birds in the study area. To 

 simplify calculations, two categories of body 

 weights were assumed, 180 g. (terns, noddies, 

 Bulwer's Petrels, and Leach's Storm Petrels), 

 and 400 g. (all others). Many species, espe- 

 cially boobies, frigatebirds, and albatrosses, 

 weighed far more than 400 g. but did not occur 

 in substantial numbers. Most of the shear- 

 waters and petrels that did occur in substantial 

 numbers weighed about 400 g. Direct migrants 

 (Sooty and Slender-billed Shearwaters) were 

 omitted from the calculations. 



Jordan (1959) studied the daily consumption 

 of the Guanay Cormorant Phalacrocorax bou - 

 gainvillii Linnaeus. He found that the average 

 Guanay weighed 2,000 g. and daily ate 430 g., 

 or 21.5 percent of its body weight. Lack (1954) 

 stated that daily food consumption of land birds 

 is tied directly to body weight, regardless of 

 the kind of food taken. The average daily con- 

 sumption of several species of land birds weigh- 

 ing between 100 and 1,000 g. was between 5 and 

 9 percent, and land birds weighing less than 

 100 g. consumed 10 or more percent of their 

 body weight daily. 



The difference in the average daily food in- 

 take of sea birds and land birds may be ex- 

 plained in part by the relatively greater amount 

 of water infishthan in seeds or insects. Taking 

 Jordan's daily consumption figure of 21.5 per- 

 cent, tern-sized sea birds weighing 180 g. 

 would eat 38.7 g. of food daily, and shearwater- 

 sized sea birds weighing 400 g. would eat 86 g. 

 of food daily. 



In one year in the study area tern-sized (180 

 g.) birds ate 3,069.7 metric tons of food in 

 79,313,614 bird-days, and shearwater-sized 

 (400 g.) birds ate 4,316.9 metric tons in 

 50,196,912 bird-days for a total of 7,386.6 

 metric tons. I assumed that birds larger than 

 shearwaters eat only as much as shearwaters, 

 so these figures are conservative. 



72 



