These two generalizations can be made from a subjective scrutiny 
of the map. Initially this survey was made with these objectives in 
mind--to find out what could be learned of dove range in the state. 
However, with the quantity of data obtained they may be useful as yearly 
indices of breeding populations to complement call counts. In order to 
compare variability of these counts with the wintering-ground mail- 
carrier indices run in the South, the data were analyzed as directed 
by Moore (1951). The results were compared with the southern analyses 
of Foote (1951), and it was found that it would be possible to detect 
nearly the same amount of population change (13% as compared with about 
10%) with less than one-third the amount of data used in the South. 
The Coefficient of Variation was also about one-third less than that 
for the much more voluminous southern data. In all probability, the 
analysis can be further refined through some sort of ecological stratifi- 
cation of the data. The data so far have only been analyzed for the state 
as a whole. Statistical comparisons between the southern and Wisconsin 
data can be seen in Table 2. 
Table 2.--Statistical Summary of Wisconsin and Southern 
Mail Carrier Surveys 
Av. No. Oo om ™ Change 
No. Doves/100 | Doves/100 | Doves/100 | Coef.] of x for 
Routes Miles Miles Miles Var. | Significance 
6 Southern 
States, } 
Winter of 1950 | 2923 16.95 29.3 0.52 173% 9.6 
Wisconsin, 
Spring of 1951 767 6.0 6.6 0.2h 110% 13% 
% Based on 39™ units giving significance at the 99.7% level. 
This greater sensitivity due to less variation doubtless is due to 
more random distribution of birds during the breeding season than on the 
wintering grounds (Nelson, et al., 1951). These counts may also be less 
vulnerable to the vagaries of weather than wintering populations, since 
Foote (1951) has pointed out that gross weather changes may greatly af- 
fect the disposition of wintering dove populations, entering serious 
sources of error into the winter inventories. 
