

14 
A-10). In South Dakota, differences were detected among 
three time periods, whereas a difference between only two 
(early and late) periods was detected in the other States. 
Thirty-nine percent of the direct recoveries of immature 
doves banded in South Dakota from 1 June through 10 July 
occurred in-State whereas 56 and 66% of birds banded 
during 11-31 July and 1-31 August, respectively, were re- 
covered in-State. These data suggest that in some States 
early-banded immatures depart natal areas before late- 
banded immatures. Thus, some early-hatched immatures 
could be banded in States other than those in which they 
were hatched. 
No chronological differences (P > 0.05) were detected 
among 10-day periods for adult doves in any State except 
for South Dakota where an increase (P < 0,05) in the pro- 
portion of in-State recoveries for 11-31 August was 
detected. This increase was based on 17 recoveries and is 
not conclusive evidence that early-banded adults depart 
natal areas differently from late-banded adults. 
These findings demonstrate a potential bias in the over- 
all analysis, particularly for doves banded as immatures. 
Since early-banded immatures may migrate earlier than 
late-banded immatures, a basic assumption underlying this 
analysis is violated. Distributional and derivational patterns 
may have been influenced by time periods in which most 
birds were banded. Those immatures that have left their 
natal areas during the banding period could be trapped, 
banded, and assumed to be part of the population in the 
capture area. This problem could be especially severe if 
most immatures were banded late in the season. Data 
related to banding success by 10-day periods were not ex- 
tracted. However, numbers of recoveries by 10-day periods 
(Table A-11) provided an indirect means of determining 
trapping success. For adult doves banded in the CMU, 
higher proportions of direct recoveries came from birds 
banded early in the preseason period (48 % in June). How- 
ever, for immatures the distribution of recoveries was 
reversed. Nearly 62% of recoveries from immatures came 
from birds banded after 20 July; 44% came from doves 
banded during August. Therefore, since a high proportion 
of recoveries of immatures came from birds banded late 
in the season, the banded sample of immature doves in this 
study may not reflect true areas of origin. 
Imprecise Indirect Band Recoveries 
When a band is reported to BBL without any reference 
as to when the bird was killed and such information can- 
not be obtained, the computer code for hunting season sur- 
vived becomes “99,” which means the year of recovery is 
unknown. Due to the computer program instructions de- 
veloped for this analysis, such recoveries were tallied as 
indirect recoveries. Since some of these recoveries are 
obviously direct recoveries (i.e., the postmark date of the 
reported band is during the first hunting season after band- 
ing), the ratios of direct to indirect recoveries are slightly 
ee 
skewed to favor indirect recoveries. About 3.4% (307) of 
the 9,091 recoveries of doves banded in the CMU were 
coded as unknown year of recovery. This aberration is espe- 
cially acute when indirect recoveries in Mexico and Cen- 
tral America are being considered since many recovery 
reports from these areas are incomplete. For instance, 
Mexico accounted for only 15% (1,386) of the total re- 
coveries (9,091), but 59% (180) of the possible erroneous 
recoveries that were included with the indirect recoveries. 
Caution should be used when comparing direct and indirect 
recovery rates, particularly when recoveries in Latin 
America are involved. 
Analytical Procedures 
Analytical procedures are discussed in the sections of the 
text where such computations were applied. Statistical dif- 
ferences are considered to be significant when P < 0.05. 
Results 
Banding Accomplishments 
Over 868,000 mourning doves were banded throughout 
the United States during the preseason banding periods be- 
tween 1967 and 1975 (Table A-12). Recoveries used in this 
analysis were from (1) doves banded in the CMU and re- 
covered anywhere and (2) doves banded in the EMU and 
WMU and recovered in the CMU and the Southern Zone. 
There were 9,822 recoveries in the selected areas of ref- 
erence (Fig. 14). Of the 447,883 bandings in the EMU, 357 
were recovered in the CMU and 52 in the Southern Zone. 
For the WMU, 88,540 bandings produced 24 recoveries in 
the CMU and 298 recoveries in the Southern Zone. In the 
CMU, 332,314 doves were banded; of these, 6,972 birds 
were recovered in the CMU, 398 in the EMU, 131 in the 
WMU, and 1,590 in the Southern Zone. These figures sup- 
port Kiel’s (1959) contention that doves originating in a 
specific management unit tend to remain in the unit. 
In the CMU, bandings were distributed fairly evenly by 
year and by State of banding (Table A-13), Yearly band- 
ings ranged from a low of about 25,000 in 1968 to a high 
of 53,000 in 1972. Total bandings by State ranged from a 
low of about 7,000 in Wyoming to a high of 48,000 in South 
Dakota. The bandings were also fairly well distributed 
geographically, although the western part of the unit was 
sampled less heavily than other areas (Fig. 15). 
Of the 332,000 CMU bandings, 182,000 were of imma- 
ture doves and 150,000 of adults. Less than 5,000 adults 
(3%) were recorded as sex unknown because of indistinct 
plumage characters. Of the 146,000 adults whose sex was 
determined, the sex ratio was heavily weighted toward 
males (187:100). Although it is now believed by some inves- 
tigators (Baskett et al. 1978; Sayre et al. 1980) that males 
