(Missouri, 83.4 vs, 68.7%: Kansas, 73.7 vs. 49.6% : Colo- 
rado, 62.9 vs. 40%; New Mexico, 78.6 vs. 56.8%; Okla- 
homa, 66.7 vs. 54.3%; North Texas, 90.4 vs. 78.6%: and 
South Texas, 95.5 vs. 78.4%, Table A-37). The area with 
a lower in-State percentage for adults than for immatures 
was Arkansas (85.4 vs. 86.2%). The composite test indi- 
cated that band recoveries of adults occurred in State of 
‘banding in higher proportions (P < 0.01) than those of 
immatures, These data suggest that immatures either begin 
their migration before adults or that they are more likely 
to migrate than adults. 
Adult males vs. adult females. — The percentage of direct 
recoveries occurring in State of banding was higher for 
males than for females in five of eight hunting areas 
examined (Missouri, 90.4 vs. 72.3%: Colorado, 64.4 vs. 
60.7%; New Mexico, 84.3 vs. 61.1%: Oklahoma, 76.7 vs. 
45.5%; and North Texas, 91.3 vs. 86.7%. Table A-38), 
Areas with lower percentages of in-State recovery for males 
than for females were Kansas, 71.5 vs. 77.5%; Arkansas, 
85.0 vs. 85.4%: and South Texas, 95.0 vs. 97.1%. The com- 
posite test indicated that band recoveries of males occurred 
in State of banding in higher proportions (P < 0.01) than 
did those for females. These data suggest that females either 
migrate earlier than males or that they are more apt to 
migrate than males. 
Summary 
Migrational patterns of mourning doves from CMU States 
of origin differ from region to region. The data suggest that 
the migrational pattern of doves is generally southward and 
shaped somewhat like an inverted fan. Doves which origi- 
nated from more northern latitudes exhibited greater prob- 
abilities of being recovered in both the EMU and WMU 
than doves originating in the southern latitudes. 
A dove banded in South Texas, Arkansas, Missouri, and 
Iowa was less likely to be recovered in Mexico and more 
likely to be recovered in the EMU than a dove banded in 
the remaining CMU States. This portion of the CMU is the 
general area described as containing an intermediate form 
of Z. m. carolinensis and Z. m. marginella (Aldrich and 
Duvall 1958), 
When distributional patterns of age and sex cohorts are 
compared, it is clear that the geographical patterns are 
similar, However, immatures are recovered out of the State 
of origin in greater proportions than adults. Similarly, adult 
females are also recovered out of State in greater propor- 
tion than adult males, This fact indicates that immatures 
cither migrate earlier than adults (and adult females earlier 
than adult males) or that they are more prone to migrate 
from State of banding. 
Distribution of the Harvest 
Band recoveries can be used to illustrate not only dispersal 
patterns of doves from specific points of origin, but also dis- 
21 
tribution of the harvest within a specific harvest area or 
areas. However, the recovery information must be weighted 
to adjust the raw data for differences in banding efforts 
among the origination areas, as well as for the varying sizes 
of the areas and density of breeding doves within them. 
Each recovery is thus placed in perspective to the number 
of birds it represents. Comparison of weighted recoveries 
for several harvest areas allows computation of the rela- 
tive importance of each. 
The weighting factors for each breeding area were 
derived by multiplying the land area value by the average 
mourning dove breeding density (as measured by call-count 
surveys) during 1967-75 (1967-74 for CMU States) and 
dividing the product by the number of doves banded during 
the same period (Kiel 1959; Dunks 1977, Table A-39). 
Breeding densities and land area values for each State were 
obtained from the 1976 Mourning Dove Status Report 
(Dolton 1977). The land area values for North and South 
Texas were taken from Dunks (1977). 
There are possible biases in using weighted data. The 
relative distribution of mourning dove harvest will be esti- 
mated incorrectly if there are differences in reporting rates 
of recovered bands among harvest areas. Likewise, if cer- 
tain States or countries do not maintain a uniform hunting 
opportunity during the period of study, the respective share 
of the harvest can be misinterpreted as an annual propor- 
tion rather than the average harvest for the duration of the 
period under consideration, Although reporting rates un- 
doubtedly vary among States (Reeves 1979), differences are 
not thought to be great. Reporting rates from areas and 
countries in Latin America are probably also relatively 
constant. It is suspected, however, that U.S. reporting rates 
may be different from those from Mexico and Central 
America (Blankenship and Reeves 1970). Consequently, we 
have chosen not to compare the relative magnitude of 
mourning dove harvests between the United States and 
Latin America. 
Distribution and Density of Harvest in the Central 
Management Unit (All Doves) 
We calculated distribution and relative densities of the 
mourning dove harvest in the CMU (Table A-40) using 
direct recoveries (Table A-45) and land area weights 
(Table A-39). 
Distribution. — Of the total harvest in the CMU, North 
Texas accounted for the highest proportion (27%), followed 
by South Texas (17%), Missouri (13%), Kansas (11%), 
Oklahoma (10%), and South Dakota and Arkansas (8 % 
each). Nebraska had closed seasons during 1967-74 and 
therefore is not represented. However, Nebraska has now 
permitted dove hunting for 7 years (1975-81) and no doubt 
presently accounts for a proportion of the harvest equal to 
or greater than that for Kansas. 
Density. — When density of the harvest is examined, the 
perception of relative importance of the harvest areas 
