
results are masked bys (1) The great fluctuations in abundance in 
different soaring areas caused by differences in wind directions (2) 
a gradual decline during late May and June in the number of albatrosses 
at Midway (apparently this decline is a function of the advancing 
season); (3) fluctuations in the daily abundance of albatrosses on and 
over the island; (4) construction work which is continually disrupting 
the normal behavior of the population. 
On May 22, the first "clean sweep" was made across experi- 
mental kill area No. 1. A total of 1495 adult albatrosses was killed. 
On the following day, runway counts at station No. 2 (Figure 21), 
which is adjacent to the killing area, revealed 246 soaring albatrosses 
per hour as compared with 96 per hour just prior to the kill. This is 
an increase of 156 percent in the birds soaring over the runway 
adjacent to the kill area. 
Runway counts, although subject to the many variables listed 
above, did not detect any consistent difference in albatrosses soaring 
over the runways after the birds had been eliminated from kill area l. 
It is realized that the population reduction experiment was conducted 
too late in the season to be most easily evaluated and was on too 
limited a scale to give really conclusive results. 
Terrain as a Factor Affecting Albatross Abundance Over Runways 
The effects of topography and wind direction on the abundance 
of albatrosses soaring over the runways were determined by counting the 
numbers of birds passing through certain index sectors. Each index 
sector was a vertical triangle at right angles to the long axis of the 
runway, delimited by the spot on which the observer stood, a point on 
the ground across the runway, and the summit of a selected tree or 
other eminence across the runway. In level areas, where no objects were 
present to measure height, the observer determined the angle of eleva- 
tion of individual birds by extending his fingers at arms length and 
judging height on a predetermined scale, so that the counts at each 
sector station are comparable. Since the apex of the triangle was on 
the side opposite the observer, a much larger amount of flying space 
was included on that side of the runway. Therefore, each count 
primarily reflects abundance on the side of the runway opposite the 
observer. Individual counts lasted ten minutes. This interval was 
established after a number of trials which indicated that under a 
given set of conditions longer counts were a waste of time. Following 
is a description of the six index counting stations at which counts 
were repeatedly made (Figure 21): 
Station 1. North end of runway 15-33. Observer on west side. 
Dunes and tall trees along east side. The trees here were taller than 
at any other point, creating the most favorable updrafts for soaring. 
Station 2. On runway 15-33 immediately south of taxiway 
intersection. Observer on west side. Area level, except for buildings 
on east side. The buildings here create small updrafts which are not 
so great as those at areas 1, 3 and 4. 
LO 
