"One-crew" Counts 
The total number of doves heard and seen on the 37 afternoon and 
37 morning counts are plotted in Figure 2. As was expected, the number 
of birds seen was much more variable than the number heard: the roadside 
count portion of the census is probably of little value except possibly 
to indicate appearance of young birds or occurrence of dispersal move- 
ments in late summer. As may be seen in Figure 2, morning call activity 
remained at a low level in March and early April, increased abruptly in 
late April, and reached a rather consistent plateau between May 10 and 
June 15, following which a gradual but persistent decline occurred. 
Afternoon counts were one-third or less as large, but approximately 
the same seasonal trend occurred. The April rise coincided with a rise 
in early morning temperatures from 0-5 to 60-65° F. Counts were quite 
consistent when run at short intervals (2-3 days) in good weather. Wind 
proved to be a very important factor; as already indicated counts were 
not made in strong wind but even at Beaufort 2-3 counts were lower, es- 
pecially early in the season. Wind apparently reduces calling activity 
and also makes it more difficult for the observer to hear doves. The 
number of calls per dove heard at 3-minute stops increased from 3.8 in 
March and early April to 6.2 during the peak of the season (May-June). 
The coefficient of variation of morning counts between May 10 and 
June 15 (peak season) is 13.8 and for afternoon counts 26.7. 
"hi ve-crew' Counts 
To determine the variation in calling activity during the tw hour 
period the number of doves heard at the first four stations visited by 
each of the five crews were totaled, then the total for the next four 
stations obtained, etc. In this manner, total doves heard by mre or 
less simultaneous observation at all stations at approximately 30 minute 
intervals was obtained. As may be seen in Figure 3, calling activity in 
the morning is very. definitely less the second hour (or 30 minutes after 
sunrise), aS compared with the first hour (or around sunrise). In the 
afternoon, however, calling did not consistently increase or decrease 
during the two hour period before sunset. This means that during a 
moming count by one crew, the last half or so of the route will be 
subject to time bias; the results obtained at the last stations will 
not be directly comparable with results obtained at stations visited 
during the first hour, or in other words, fewer doves will be heard at 
the end of route even if the population is the same as at the beginning. 
In Figure h, the time bias is removed (5-crew count data) and each of 20 
stations evaluated as to morning calling activity. The index used is the 
total number of doves heard at the station by all five crews multiplied 
by the number of crews (1-5) recording doves at the station. During 
April, May, June and July at least one dove was heard by at least one 
crew at every station while all crews recorded doves at the more active 
stations. Figure lh, indicates that there were no pronounced shifts in 
