
10 
ports). The annual United States quota was set at about 
175,000 and each State was to contribute about 3,000-5,000 
bandings each year (smaller eastern States had quotas of 
only 250 bandings and California and Texas had highs of 
8,000). Instructions to State agencies stressed the need for 
wide distribution of banding effort according to physio- 
graphic stratification throughout each State. No more than 
500 birds were to be banded in any one area. 
The CMU banding effort was coordinated through the 
Central Migratory Shore and Upland Game Bird Techni- 
cal Committee, an auxiliary of the Central Flyway Council. 
The banding effort was cooperatively undertaken by both 
State and Federal banders; some private banders partici- 
pated. Much of the banding effort by State fish and wild- 
life agencies was supported by contracts issued under the 
Accelerated Research Program for Migratory Shore and Up- 
land Game Birds (Sandfort 1977:2) and by Federal Aid to 
Wildlife Restoration. 
Objectives of Banding and 
Underlying Assumptions 
Banding, the marking of animals to allow identification 
of individuals, provides a means for obtaining data about 
the characteristics and dynamics of a population. 
- The direct recovery rate provides an index to shooting 
pressure. 
- The recovery rate, when adjusted by the reporting rate, 
provides an estimate of the harvest rate, or the proportion of 
the banded cohort which is harvested and retrieved by 
hunters. 
* The harvest rate, when adjusted by crippling loss, pro- 
vides an estimate of the overall kill rate. 
* The distribution of recoveries indicates location and 
relative harvests by geographical areas and time periods for 
specific banding areas. 
* The derivation of recoveries depicts the banding origin 
of birds harvested in specific recovery areas. When weighted 
by population information, the relative importance or con- 
tribution of each banding location to a recovery location 
can be calculated. 
* The band recovery pattern, although emphasizing har- 
vest aspects, is an indirect indicator of geographical and 
temporal migratory patterns. 
- Estimates of annual survival can be calculated from 
the number of birds banded and the temporal distribution 
by year of subsequent recoveries. 
: Indirect estimates of population size existing at the time 
of banding may be calculated from band recovery, band 
reporting rate, and harvest data. 
Inasmuch as banding is a sampling technique, it is impor- 
tant that the marked or banded sample be representative 
of the overall population being studied. To be representa- 
tive, birds must be banded uniformly throughout their 
range, by time periods, and according to age and sex. For 
the CMU mourning dove banding program, planners 
sought to distribute the banded sample evenly throughout 
the doves’ range, to capture birds during the breeding sea- 
son so as to establish their points of origin, and to obtain 
adequate samples of each age and sex class. These objec- 
tives were basically met by following guidelines established 
by the cooperative banding program described earlier. 
However, because of monetary and time restraints, banders 
were not always able to evenly distribute bandings geo- 
graphically, chronologically, and by age and sex classes. 
Therefore, some of the conclusions presented herein for 
some States are biased toward certain areas of origin, time 
periods, or age and sex cohorts depending on trapping 
efforts within those States. 
To reduce trapping effort bias among banding States, 
the recovery data were weighted by population and band- 
ing statistics when calculating derivation of the harvest. 
This practice not only corrected for disproportionate sam- 
pling of population components but also for differences in 
the sizes of the sampled populations and allowed compari- 
son of the relative contributions of originating populations 
in specific harvest areas. 
Certain assumptions which may or may not be true have 
to be made when analyzing banding data. In this analysis, 
the following assumptions were made: 
- That bands placed on doves were retained by them 
throughout their life span. This assumption is probably 
correct. 
* That birds banded during June-August represent 
breeding adults or young reared in the vicinity of capture. 
Available evidence indicates that this assumption is not 
entirely correct; some birds may be migrating during this 
period and therefore trapped away from their site of origin. 
* That the various age and sex cohorts are trapped 
(sampled) in the same proportion as that of the population. 
This assumption is important when combining data because 
if one age or sex component is sampled disproportionately, 
the combined data could mislead the analyst about char- 
acteristics of the entire population. 
° That banding samples are large enough to minimize 
the importance of sampling error. For computations on a 
Statewide or larger basis, this assumption is probably met 
for this analysis. 
- That reporting rate of recovered bands is uniform 
among States in the United States. This assumption is incor- 
rect to an unknown degree. Two studies (Tomlinson 1968; 
Reeves 1979) shed some light on this problem, but their data 
are insufficient to allow assignment of reporting rate values 
by State of recovery. Therefore, an average figure is arbi- 
trarily used herein. Reporting rate for recoveries in Mexico 
was estimated according to Reeves’ (1979) data. 
* That doves banded in different areas of a particular 
State have the same migrational tendencies. This assump- 
tion is also basically incorrect because doves are influenced 
