estimating technique. However, for most of 
the wintering areas it is believed that the 
yearly data for geese and certain other spe- 
cies such as the black duck and swan can be 
used to indicate broad trends in population 
size, In addition, the winter survey data have 
great value for determining the use made of 
various wintering areas and the change in 
waterfowl distribution from year to year in 
response to changing weather and habitat 
conditions, 
All data referring tothe 1963 winter survey 
are based upon observed birds, 
Section C (pp. .79-103) of the appendix con- 
tains tables summarizing the results of the 
annual winter survey. 
BREEDING GROUND SURVEY 
Data supplied by Fred A, Glover 
Bureau of Sport Fisheries and Wildlife 
Surveys are conducted each year onthe 
waterfowl breeding grounds for the purpose 
of estimating the relative size of the fall 
flight from each of the breeding areas, Two 
coverages of the breeding areas are required 
to obtain the necessary information: the first, 
during May and June, to measure distribution 
and relative size of the breeding population; 
the second, during July, to forecast the rela- 
tive number of young that would be produced, 
In July it is necessary to makea preliminary 
estimate or forecast of the number of young 
that will be produced, since only a part of 
the séwagn’s young will have hatched at the 
time field work must be terminated so that 
data will be available for setting the shooting 
regulations. The production survey, there- 
fore, consists of a measure of the number of 
broods on the water at the time ofthe survey 
plus a measure of weather, water, and other 
conditions that affect or reflect production 
success following the survey period. 
Most of the importarit waterfowl breeding 
areas in Alaska, Canada, North Dakota, South 
Dakota, and Minnesota are surveyed fromthe 
air. Statistically designed sampling techni- 
ques and similar methods of collecting and 
analyzing data are used throughout these 
areas. In addition to the areas mentioned, 
many of the Northern States conduct breeding 
grounds surveys. Methods vary somewhat 
among these States, although in States with 
important numbers of breeding ducks the 
methods are similar in most respects to 
those employed in the Dakotas, Canada, and 
Alaska, 
In recent years, aerial crews have sampled 
approximately 2,375,000 square miles of the 
best duck breeding habitat on the North 
American Continent. The only important 
duck breeding areas that are not being cen- 
sused by standard survey procedures are 
those in eastern Ontario, Quebec, and Labra- 
dor, for which adequate census techniques 
and data analysis have not yet been developed. 
Experimental surveys are being conducted in 
this area but they have not progressed to the 
point where reliance can be placed on the 
findings for forecast purposes. 
The aerial crews count the birds on some- 
what less than 1 percent of the total breeding 
area. This is sufficient coverage to reduce 
Sampling error to less than 20 percent ofthe 
average population density in most survey 
areas, and to much less than 20 percent when 
considering the breeding range as a whole. 
The results of the breeding ground surveys 
are presented as indexes, When conducting 
aerial surveys of breeding birds or of broods, 
not all birds present are seen by the aerial 
crews. Methods of measuring the proportion 
of birds present that are seen are being 
developed, but these studies have not pro- 
gressed to the point where visibility factors 
can be determined throughout the breeding 
range. Since there is no attempt to estimate 
the number of birds not seen, the indexes 
presented in this report are based on birds 
actually seen, and it is emphasized that they 
do not constitute estimates of the total num- 
bers present. 
Results of the May survey of the breeding 
population and of the later production survey, 
when combined, form the basis for forecasts 
of changes in the relative size of the fall 
flight of ducks and coots in the three Flyways 
from the Mississippi Flyward westward. It 
is not possible to rely on the breeding ground 
information to the same degree inthe Atlantic 
Flyway as in the other Flyways, primarily 
because of the lack of adequate survey data 
from Quebec and Labrador, which are im- 
portant contributors of birds to that Flyway. 
Experimental aerial surveys were conducted 
