18 

Fig. 16. Relative frequencies of recovery in the Eastern Manage- 
ment Unit for adult mourning doves banded in designated areas 
of the Central Management Unit. Example: An individual dove 
banded in Minnesota, Iowa, Missouri, or Arkansas is 56 times 
more likely to be recovered in the EMU than a dove banded 
in any of the remaining CMU States (56/1 = 56). 
probabilities of being recovered in a selected harvest area, 
we tested the recovery data for independence using the 
G-test (Sokal and Rohlf 1969). The data were separated 
for doves banded as adults (Table A-17) and immatures 
(Table A-14) in each of the 15 CMU banding States. Three 
discrete harvest areas (EMU, WMU, and Mexico) were 
examined. Previous testing indicated that recovery rates for 
adults were different from those for immatures; thus these 
data were not pooled. 
For each of the 15 banding locations, the number of doves 
banded in relation to the number of doves recovered in a 
specified harvest area (the EMU, WMU, or Mexico) were 
used to test for independence. When differences were 
detected among banding States, the CMU was partitioned 
by inspection (a posteriori) into groups of States with similar 
dove recovery patterns. The pooled data for each parti- 
tioned area then were tested within and among areas. When 
these tests confirmed the integrity of the areas, combined 
recovery rate indices were computed for each. We then cal- 
culated comparative recovery factors using the banding area 
with the lowest recovery rate index as a factor of 1. The 
CRF values that were obtained reflect the relative frequen- 
cies at which doves from the banding areas were recovered 
in the specified harvest locations (Figs. 16-21), 
Tendency for adult doves to migrate to the Eastern 
Management Unit. — The tests for independence of doves 
banded as adults and recovered in the EMU indicated two 
Fig. 17, Relative frequencies of recovery in the Eastern Manage- 
ment Unit for immature mourning doves banded in designated 
areas of the Central Management Unit. Example: An individual 
dove banded in Missouri is twice as likely to be recovered in the 
EMU as a dove banded in Minnesota, Iowa, and Arkansas 
(91/43 = 2.1). 
groups of banding areas that were similar within each group 
but different between the groups. The recovery rate index 
in the EMU for the eastern tier of CMU States (Minnesota, 
Iowa, Missouri, and Arkansas) was 14.48; and for the re- 
maining CMU States (Montana, North Dakota, Wyoming, 
South Dakota, Nebraska, Colorado, Kansas, New Mexico, 
Oklahoma, North Texas, and South Texas), it was 0.26. 
The CRF’s are 56 and 1, respectively, and therefore the odds 
were 56 to | that an adult dove from the eastern tier of 
CMU States would be recovered in the EMU over an adult 
dove from the remaining CMU States (Fig. 16). 
Tendency for immature doves to migrate to the Eastern 
Management Unit.—The grouping of CMU States with 
doves demonstrating similar migrational tendencies to the 
EMU is more diverse when using data for doves banded 
as immatures. G-test results indicate four areas with distinct 
recovery rate indices (Fig. 17). The highest recovery rate 
index in the EMU was for immature doves banded in 
Missouri (67.32). For Minnesota, Iowa, and Arkansas, the 
recovery rate index in the EMU was 31.55: and for North 
Dakota, South Dakota, and South Texas, it was 7.96. The: 
lowest recovery rate index in the EMU (0.74) was from 
doves banded in the remaining CMU States (Montana, 
Wyoming, Nebraska, Kansas, Colorado, New Mexico, 
Oklahoma, and North Texas). Thus, an immature dove 
banded in Missouri had the highest CRF (91); followed by 
Minnesota, Iowa, and Arkansas (43); North Dakota, South 
