If this correction is applied to the data for P. sexta males given in 
fisure 2, the regression equation then becomes log y = 2.003 - 0.266 X, and the 
regression coefficient is highly significant. By substituting appropriate 
values in this equation, it can be calculated that 5 percent of a dispersing 
population will travel }.89 miles. 
Since a correction has been applied for the thinning effect of dispersal 
outward from a point, we can now say that in a circle with a radius of 4.89 
miles, 5 percent of the moths at the center will have come from outside the 
circle. Since a certain minimum area is needed to measure a population, an 
experiment was designed in which the area trapped had a radius of 6 miles. It 
would be expected that somewhat less than 5 percent of the moths inside a center 
circle with a radius of 1 mile would come from outside the trapped area. 
After estimating the increase in total catch to be expected from increasing 
the number of traps, it was concluded that three traps per square mile were 
necessary to control the population of P. sexta. Since there are a little over 
113 square miles in a circle with a radius of 6 miles, the number of traps 
required would be 339, but in practice large areas of forest reduced the number 
to 324. This experiment was put in operation in May 1962. Not all the traps 
could be checked each day, and after some preliminary experiments 28 traps out- 
side the trapped area and 24 inside were selected. These were examined each 
day from July 14 to October 10. 
Figure 4 shows the relationship between the distance of each trap from the 
center of the trapped area and the total catch per trap of P. sexta. Curves 
for P. quinquemaculata were very similar. The catch per trap of both species 
and both sexes cradually increased as distance from the center increased, and in 
no case was there a sharp rise in catch at the edge of the trapped area. Thus, 
hornworm movement tended to obliterate differences between populations inside 
and outside the trapped area. 
As a result, a simple comparison of mean catch inside and outside will 
underestimate the effect of the traps in reducing the population. A better 
estimate can be obtained by calculating the catch at the center and 12 miles 
out, that is, at equal distances from the border of the trapped area. These 
estimates are given in table 7. 
TABLE 7.--Estimated reduction in populations of hornworm moths due to 
3 light traps per square mile, July 1hy=-Oct. 10, 1962 
Calculated catch at-- Percent 
pecies and sex 12 miles reduction 
