Breeding Population and Production Surveys 
Surveys are conducted annually on the waterfowl breeding 
grounds for the purpose of estimating changes in the relative size 
of the fall flight from each of the breeding areas. Two surveys 
of the breeding areas are required to obtain the necessary informa- 
tion. The first survey is made during May and June for the 
purpose of measuring the distribution and relative size of the 
breeding population. The second survey is carried out during July 
for the purpose of forecasting the relative number of young pro- 
duced. In July it is necessary to in part "forecast" the number 
of young that will be produced since by the time field work must 
be terminated in order to have the data available for use in 
setting shooting regulations, only a portion of the total season's 
production will have hatched. The production survey, therefore, 
consists of a measure of the number of broods on the water at the 
time of the survey plus a measure of weather, water, and other con- 
ditions which affect or reflect production success after the survey 
period. 
The bulk of the important waterfowl breeding areas in 
Alaska, Canada, North Dakota, South Dakota, and Minnesota are 
surveyed from the air using statistically designed sampling tech- 
niques. Similar methods of collecting and analyzing data are 
used throughout these areas. In addition to the areas mentioned 
above, approximately 22 of the Northern States conduct breeding 
ground surveys. Methods of conducting surveys vary somewhat among 
these States, although the methods employed in States with 
important numbers of breeding ducks are similar in most respects 
to those employed in the Dakotas, Canada, and Alaska. 
During recent years aerial crews have sampled approxi- 
mately 2,375,000 square miles of the best duck breeding habitat on 
the Continent. The only important duck breeding areas that have 
not been censused are those in Eastern Ontario, Quebec, and 
Laborador. Surveys have not been conducted in these areas due to 
the fact that adequate census techniques have not yet been 
developed. 
The aerial crews count the birds on somewhat less than 
one percent of the total breeding area. This amount of flying is 
sufficient to reduce sampling error to less than 20 percent of 
the average population density in most survey areas, and to much 
less than 20 percent when considering the breeding range as a 
whole. 
