stayed in areas away from the tidewater. 
Further discussion of aerial-survey methods 
and their limitations is given by Stewart, Geis, 
and Evans (1958), 
Additional information concerning distri- 
bution and numbers of waterfowl at different 
seasons came from numerous local counts 
made from land or by boat. All important 
tidewater areas of the Upper Chesapeake 
region were covered in these ground surveys. 
Data were particularly useful for some of the 
more inconspicuous species that were difficult 
to census from the air. Most of the counts 
on open-water areas of the coastal and 
estuarine bays were made with the aid of a 
20-power telescope from systematically se- 
lected observation points. All waterfowl within 
a radius of a quarter ofa mile could be counted 
by this method. Waterfowl in marsh areas 
were counted from a motorboat along measured 
courses of tidal creeks or guts; with the aid 
of 7-power binoculars, all waterfowl observed 
within an eighth of a mile were counted. River 
marshes and many bay marshes could be 
completely covered from the watercourses, 
but in a few of the larger bay marshes it was 
necessary to make separate counts in various 
marsh ponds or "broken marshes" farther 
away from the watercourses. Counts from 
different years and different trips were com- 
bined by seasons to give a composite picture. 
Information concerning plant associations 
was obtained from extensive ground surveys. 
Each of the major waterfowl areas was covered 
systematically. Lists of plants were made for 
numerous stations in each area, as well as 
detailed notes on abundance, associated plants, 
salinities, tides, and other environmental fac- 
tors. Type maps were made for the entire 
area. Information on occurrence of inverte- 
brate animals was obtained during the ground 
surveys and from the food contents of water- 
fowl. 
Food-habits information was obtained from 
field observations and from examination of 
gullet and gizzard contents in the laboratory. 
Most of the specimens for food analysis were 
collected by the author, but considerable 
numbers also were collected by various per- 
sonnel of the U. S. Bureau of Sport Fisheries 
and Wildlife and the Maryland Game and 
Inland Fish Commission during and before the 
present study. Analyses were made of 1,240 
specimens. Of these, 952 were examined by 
the author, and 288 were examined by other 
biologists, chiefly F. M, Uhler and Clarence 
C. Cottam, The volume of the entire contents 
of the gullet and gizzard of each specimen 
was measured. The volumetric proportion 
(percent) of each food item was estimated 
visually. 
The food-habits data were organized ac- 
cording to the habitat types from which the 
birds were collected. Food-habits data were 
expressed as percentage of birds in which the 
various food items occurred. Any occurrence 
making up less than 5% of the particular 
stomach contents was omitted from the tabula- 
tion. In order to express the reliability of the 
percentage-occurrence figures, 95-percent 
confidence limits were read from the tables 
of Mainland, Herrera, and Sutcliffe (1956), 
Information concerning the local distribution 
of the kill for a given species was based on 
the percentage distribution of band recoveries 
from birds reported as -shot in the Upper 
Chesapeake region. Only recoveries of birds 
that had been banded outside Maryland and 
Virginia were used in these computations, 
and the data were adjusted so that each year's 
recoveries had the same weight. This was 
done by considering the total weight of all 
recoveries to be 100% and by calculating the 
weighted value of recoveries for each year as 
100% divided by the number of years repre- 
sented by recoveries. The weighted value of 
a single recovery for any particular year 
could then be obtained by dividing the weighted 
value of recoveries for each year by the 
number of recoveries for that particular year. 
Information concerning distribution of the 
kill of Chesapeake birds throughout North 
America was obtained from birds banded in 
the Upper Chesapeake region. Conclusions 
were based on the percentage distribution 
of indirect recoveries from birds reported 
as shot, 
Data on seasonal occurrence, migration 
peaks, and arrival and departure dates were 
taken chiefly from ''Birds of Maryland and the 
District of Columbia’! (Stewart and Robbins, 
1958), 
Identities and Latin names of all plants 
except algae are according to the eighth 
edition of Gray's manual (Fernald, 1950), 
Common and scientific names are listed inthe 
appendix. Mollusk names were based chiefly 
on ''A Field Guide to the Shells" (Morris, 1951). 
Most of the names of other invertebrates were 
taken from ''Field Book of Seashore Life'' 
(Miner, 1950), and''Wardand Whipple's Fresh- 
water Biology'' (Edmondson, 1959), Scientific 
names of fish and invertebrates were brought 
up to date by authorities at the U. S, National 
Museum, U, S, Agricultural Research Service, 
and the U. §S. Fish and Wildlife Service. 
The field and laboratory investigations were 
conducted with the halftime assistance of 
Don P, Fankhauser over a 2-1/2 year period 
and with the help of Charles F, Kaczynski for 
three-fourths of his time during 1 year. Both 
men worked with diligent perseverance toward 
the completion of a variety of tasks, including 
some that were physically strenuous and 
others that were very tedious. 
The generous cooperation of many biologists, 
game technicians, and law-enforcement per- 
sonnel is deeply appreciated. Charles D, Evans, 
Atlantic Flyway Biologist, was primarily re- 
sponsible for the periodic aerial surveys of 
waterfowl populations during the 1958-59 sea- 
son. Vernon D, Stotts and Richard N, Smith of 
