Large samples of unemployed birds were marked with Krylon 
enamel paint, readily distinguished from the dyed nesting individuals. 
The locations of painting, the numbers painted, and reference symbols 
for each group are shown in Figure 15. In addition to those shown on 
the map, 400 unemployed birds designated by the symbol "NG" were dyed 
in March within area "B®", The Eastern Island birds were marked on 
March 73 the others, between January 22 and February 14. 
Half-hour and one-hour counts of birds flying over the run- 
ways were made at various times of the day during the period December 
2? through March 18 for nesting birds, and February 14 through March 
18 for unemployed birds. Sixty-eight hours were spent in malcing these 
observations through March 18, and the counts were divided about 
equally among the locations shown by numbers 1, 2, 3, and 4 on the 
Send Island runways (Figures 14 and 15). 
Analysis 
Statistical tests were conducted to test variation in per- 
centage occurrence of the various marked birds, and correlation if any 
with time of the day, period of the winter, and observation location. 
Field observations indicated, and statistical tests confirmed that the 
blue dye (Brilliant Milling Blue) faded so rapidly as to be of no use 
for the present study. For purposes of analysis the birds in the 100- 
foot wide blue strips were allocated to the adjacent violet areas and 
considered as unmarked violet birds. 
Construction operations interfered considerably with field 
investigations. in particular, a large area that contained violet 
breeding birds was bulldozed, destroying the nests and transferring 
an unknown number of violet birds to unemployed status. There was a 
highly significant increase in violet birds over the runway on and 
after February 21 (the day of greatest disturbance). A corresponding 
highly significant decrease in proportion of green birds occurred at the 
sane time; this was probably due in part to appearance over the runways 
of many newly unemployed birds from the violet area, and possibly a 
result of other factors not immediately apparent. 
In view of the drastic changes in percentage occurrence of 
some segments of the population from February 21 on, related in large 
part to habitat disturbance, only those observations prior to this date 
are included in the analysis in Table 5. Of 10,155 observations of 
Laysan albatrosses over the runways during this period, 166 or 1.6. 
percent were dyed breeding birds. Knowing the percentage of birds 
dyed in each area, an estimate of the total number of breeding birds 
represented in the runway counts was computed. This number, 4722, 
comprises nearly half of the total runway observations and indicates 
that breeding birds contribute about half of the birds over the run- 
ways, and unemployed birds the other half. 
28 
