32 
stant was tested with data from Models | and 2 of program 
ESTIMATE, Although the power of this test is low and appro- 
priate only if Model | fits (P > 0.05), four of nine refer- 
ence areas tested indicated a rejection (P < 0.05) of the 
hypothesis (Table A-79). The composite test also indicated 
a rejection; therefore, annual survival rates were not con- 
stant. 
Life Equation 
Mathematical models relating mortality rates to recruit- 
ment rates necessary for maintaining stable populations are 
referred to as life equations and have been described by 
Henny et al. (1970). Using survival estimates for adult and 
immature doves (Tables A-73 and A-74), we computed the 
amount of production needed to compensate for annual 
mortality to maintain a constant breeding population for 
each State (Table A-80). The production requirement 
ranged from 1.2 (Nebraska) to 4.4 (Arkansas) young per 
pair of breeding adults. The production requirement in the 
entire CMU was 2.2 young per pair of breeding adults. 
A compilation of 17 mourning dove nesting studies across 
the United States (U.S. Department of the Interior 1977) 
indicated the average number of fledglings produced per 
pair of adults was 3.3 in northern latitude States, 4.4 in 
mid-latitude States, and 3.7 in southern latitude States. The 
overall average was 4.0 fledglings per pair of breeding 
adults. The estimated amount of production required to 
maintain a constant breeding population seems well within 
the measured capabilities reported in the literature. 
Differential Vulnerability 
Immature mourning doves consistently had higher direct 
recovery rates than adults. The difference between these 
rates is attributed to differential vulnerability between the 
age classes. Relative recovery rates were calculated for each 
State in the CMU and averaged for the unit (Table A-81). 
The relative recovery rate varied from 0.87 for Kansas to 
2.90 for Montana among CMU States, but the high rates 
for Montana and Wyoming were probably due to low 
sample sizes, Excluding data from those two States, the 
average relative recovery rate for the CMU was about 1.3. 
Thus, immatures were slightly more vulnerable to hunter 
harvest than adults. This value can be used to adjust age 
ratios obtained from hunting season wing collection sur- 
veys to reflect the true preseason proportions of young to 
adult mourning doves in the CMU. 
Proportion of Population Removed by Hunting 
Each year, hunting removes a large number of mourn- 
ing doves from the population. Ruos and Tomlinson (1968) 
estimated that 41 million mourning doves were harvested 
in the United States in 1965. Keeler (1977) compiled data 
from various sources and estimated about 49 million birds 
killed in 1972. 
The proportion of the CMU dove population which died 
during the 1967-74 period due to hunting was calculated 
by using the method described by Geis (1972). Kill rates 
for adults and immatures were calculated by adjusting the 
recovery rates (Tables A-73 and A-74) for adults and imma- 
tures, A reporting rate of 45% (Reeves 1979) and a crip- 
pling rate of 30% (Winston 1954; Dahlgren 1955; Shaw 
1956; Nelson 1957; Haas 1977) were used as the adjusting 
factors. The kill rates for adult and immature doves from 
the CMU were 4.7 and 6.3%, respectively. Therefore, the 
estimated proportion of total mortality (47.3% for adults 
and 56.2% for immatures) due to hunting was 9.9 and 
11.2%, respectively (Table A-82), 
Anderson and Burnham (1976) reported the effect that 
hunting had on mallard (Anas platyrhynchos) populations 
and tested the concept of hunting being completely com- 
pensatory to other forms of annual mortality. They re- 
viewed the literature and found most conclusions regarding 
hunting mortality as additive mortality were based on 
incorrect and inappropriate methods. They rejected the 
concept of hunting being completely additive to other mal- 
lard mortality but could not reject the concept of hunting 
being completely compensatory. In the mallard study, 
40-50% of the birds dying in any given year were killed 
by hunters (Anderson 1975). This mortalitv was largely, 
if not completely, compensatory. Although the kill rate level 
which will adversely affect mourning dove populations is 
unknown, it is difficult to conceive that hunting has 
adversely affected a mourning dove population that sus- 
tains a hunting mortality that is only 10-11% of the total 
mortality. 
Indirect Population Estimates 
The abundance and distribution of the mourning dove 
breeding population is annually described in the form of 
an index (Dolton 1977), For evaluating trends in the popu- 
lation, this method is sufficient. However, indices are dif- 
ficult to compare with other estimates such as harvest and 
disease losses which are expressed as absolute numbers, For 
this reason, data derived from this banding analysis were 
coupled with harvest data to estimate, numerically, the 
CMU and nationwide mourning dove fall flight popula- 
tions, These estimates are founded on broad assumptions 
and should be viewed cautiously. 
The basis for this computation is to derive two essentially 
independent estimates of the kill of mourning doves from 
a defined geographical area. One estimate employs banding 
data to project the proportion of the total population killed 
by hunters. The other estimate employs harvest and band- 
ing data to project the absolute value of this kill. By divid- 
ing the second value by the former, an absolute value for 
the total population is obtained. This procedure was dis- 
cussed by Geis (1972) and used by Dunks (1977) in project- 
ing the Texas mourning dove population. 
The fall flight estimates are strongly dependent on an 
estimate of the band reporting rate. The 32% band report- 
ing rate (Tomlinson 1968) was used rather than the 45% 
