These surveys indicate an appreciable increase in potential nesting water- 
fowl in the Lake Erie Marshes, a substantial increase in the farmland pothole 
country of Barry County and also a slight increass in the transsct through the 
multi-lake country of Ingham, Livingston, ena Cakland counties. 
The over-all change from 1.39+ local breeders to 3.384 per square miles, a 
143 percent increase, is sncouraging. 
In order to chock on ths accuracy of aerial censuses, a comparison was made 
with a ground survey in four townships in Barry Vounty. A Block-type aerial 
survey was. made of these townships to obtain completes coverage. At the same time, 
other Department personnel carried on a ground survey by checking all water and 
marsh habitat in 24 section, or 16.66 percent of the 144 square mile arsa censused 
by airplene. Ths results wers similar. Ths asrial survey showsd an average of 
.80 nesting birds per square mile whils the ground check gave .75 potential nest- 
ing ducks per square mile. This ground check supports the findings of our block 
type of asrial survey in the farm pond type of habitat in southern Michigan. 
Aerial and Surface Censuses 
While the Barry County check worked out quite satisfactorily, our asrial 
census figures elsewhere were not comparable to surface censuses mads by boat. 
The area that could be covered by boat was, of course, greatly limited in compari- 
son to that coversd by plans. The boat censuses also represented the best habitat 
along water courses where birds were likely to concentrate. In order to illustrates 
this point the results of both types of surveys are given for one marsh. Following 
an aerial census a strip census by boat was made on the 2,600 acre Pte. Mouilles 
Marsh adjacent to Lake Erie. Complete coverage by air showed 11 ducks per square 
mile, while our surface check of a portion of the same marsh gave 8&8 potential 
breeding birds per square mils. Obviously, the data obtained by these two survey 
methods are not comparable. Both methods are satisfactory, in themselves, when 
used as relative indexes comparing results from one year to another. They cannot 
be compared to sach other, for one represents an sxtensive survey of all types of 
habitat, and the other an intensive coverage of good habitat. 
Surface Censuses 
Ground checks were made this year for the first time on sample areas to obtain 
greater coverage to provide for comparative indexes. Our District Game Managers 
censused 13 areas. A count of 585 potential local nesting ducks was obtained on 
85 linear miles of census line, an average of 6.8 ducks per linear mils. The width 
of the census strip varied from one hundred to several hundred feet giving a total 
of only 10.36 square miles. The computed breeding population averaged 56 ducks 
per square mile. 
The composition of species was 45.5 percent black ducks, 35.2 percent mallards, 
12.9 percent blue-winged tsal and 4.3 percent wood ducks. Other species nesting 
in small numbers made up the remaining 3.1 percent. 
There were 61 ducks observed of species that rarely, if ever, nest locally. 
Thess individuals were not counted in the above totals, Details of this survey 
by sample areas are shown in Table l. 
195 
