
Taner | a) 
gives an average figure of 0.98 breeding pairs per square mile for those areas 
covered as compared to a ground census firure of 0.76 pairs per square mile. 
In the opinion of the writer, this discrepancy is largely attributable to the 
fact that many more large water arsas are covered more completely by aeriel 
census. Consequently the number of summering drakes counted on aerial transscts 
was appreciably highsr than G6bserved on ground transects because such summering 
birds tend to flock on much larger water areas, This point is woll demonstrated 
by data obtained from the lake and river surveys and is particularly evident in 
the case of black ducks and mallards. 
Strangely, the mean average of both systems of computing breeding pairs 
per square mile is 0.705 breeding pairs per square mile for both the aerial 
and ground data. In the opinion of the writer the actual breeding population 
is in excess of one breeding pair per section, in light of the fact that many 
birds are obviously missed by aerial coverage. However, since the purpose of this 
census system is primarily to measure trends from year to ysar ths point may bs 
irrelevant. 
Laks and Marsh Surveys 
Forty Wisconsin lakes and marshes were surveysd by flying thse shorslins at 
a distance of 100 to 200 yards. An effort was made to count all birds along a 
standard 220 yard strip on either side of the plane so that such data could be 
expressed in terms of birds and pairs per square mile. (To obtain birds and 
estimated pairs per lineal mile of shoreline the data per square mils should 
bs divided by four.) Admittedly, area covered may not be computed as accurately 
as on ground and cross-country aerial transects; but we have attempted to reduce 
laks, marsh, and river surveys to the same common denominator as employsd in 
regular transects. Such surveys should be fully indicative of trends from year 
to year as standard transscts; moreover, guch surveys have the advantage of 
dsaling with larger numbers of waterfowl which should tend to reduce statistical 
errors. 
Lakes covered consisted of 21 lakes and marshes in northwestern Wisconsin 
in Burnett County and 19 principal lakes and marshss in southeastern Wisconsin 
in the Fox and Wolf River valleys. These surveys entsiled a total of 77 square 
miles of water and marshland area on which over 1,600 ducks and over 500 coots 
were counted. An average breeding density of 3 to 7 pairs of waterfowl was 
found with densities in excess of 20 pairs per square mile recorded for a few 
areas. 
These surveys also definitely indicate that WPisconsin marshes do produce 
substantial numbers of waterfowl. Data obtained is as follows: 
189 
