190 
Table IV - Comparison by Species of Air-to-Ground Observations of Ducks 
Se SSS 
Percent of Ground-Counted Birds Seen from the Air 
a 
peak Mallard-like Teals 
1951 55 8 
1952 64 59 

Brood Survey 
The aerial brood count study for both 1951 and 1952 revealed that the 
air-to-ground ratio was not consistent between coverages. This was checked with 
the chi-square test resulting in a calculated chi-square value larger than the tabular 
value of chi-square .5. Therefore, it is acknowledged that aerial brood-counts on 
irrigated land in the Cache la Poudre Valley cannot be corrected to compare with 
ground counts. Neither may these data be used for year-to-year trend purposes, as 
fluctuation in the number of broods counted per coverage was not necessarily caused 
by more or less broods being present on the ground. Preliminary studies on other 
land types in Colorado, however, definitely indicate that correction factors may be 
applied to aerial brood-counts. 
Table V compares air and ground estimates of average brood sizes for 
1951 and 1952 showing that the differences were slight for both years. The smaller 
deviation between air and ground estimates in 1952 was attributed to the more 
experienced observer. 
Table V - Comparison of Air and Ground Average Brood Sizes 

1951 1952 
Air Ground * Air Ground * 
Number of Broods 21 59 27 38 
Number of Young 112 348 149 221 
Average Brood Size 5.33 | 5.90 5.52 5.81 
* Only accurately counted broods are included from comparable ground 
counts. 
Conclusions 
From the foregoing analysis and discussion, it is possible to draw the 
following general conclusions: © 
1. A trained observer is essential to the success of aerial waterfowl 
census such as described. 
