
and broods could be expected to show themselves if not disturbed during the 
waiting period. No attempt was made to approach the pothole without distur- 
bing the birds, as this was found to be impossible. If the observer then 
concealed himself or remained absolutely still, activity generally was re- 
sumed within 30 minutes of his arrival. Depending on water levels, vegetation, 
and available personnel, from 20 to 45 potholes were censused in this manner 
on each coverage of the study area. These contained over three-fourths of the 
broods seen. 
In 1951 and 1952, a Labrador retriever was used to assist in finding 
broods. The dog was generally kept under close control during the zigzag 
"beat-out"” and was allowed freedom at the end to flush, but not to chase the 
birds. He was later sent up on shore to make sure that none had escaped over- 
land. These he was allowed to chase as they were difficult to find otherwise. 
The dog was almost a necessity in 1951 and 1952, since water levels were high 
and many of the overgrown areas contained water throughout the brood season. 
In 1950, most overgrown areas were dry, or nearly so, before many broods had 
hatched. Open-water areas in this region could be well covered without the 
dog, provided care was taken to discourage broods from going ashore. 
Brood census was not attempted in winds much over 15 miles an hour, be- 
cause of the difficulty in driving the birds. 
In general, it is believed that the coverage for broods was nearly com- 
plete, but it should be noted that the production figures for the study area 
represent a minimum. 
Late-summer and fall counts of adults and flying juveniles were also made 
in conjunction with brood counts. Since these were made mainly for the 
purpose of determining utilization, only rough estimates were made of popu- 
lations of flying birds, and no attempt was made to separate adults from 
juveniles. Thirty complete coverages of the study area have been made: eight 
in 1950, ten in 1951, six in 1952, and six in 1953. Twelve additional 
coverages of a 2-square-mile area were made in the summer of 1950, and five 
extra coverages were made of a 1l-square-mile area in the spring of 1951. 
Other data 
“Records were kept throughout the season of each year on the depth of each 
pothole whenever it was visited. 
During the late summer and fall of 1950, rough cover maps were made of 
each pothole. In subsequent years, notes were taken on the condition of the 
vegetation at the time of each visit. These were simple and included a rough 
estimate of the width of each band of vegetation, its predominant species, and 
density. 
Using the data on water levels and vegetation, it was possible to modify 
the original classification developed by Bach (1952), which depended on 
permanence, to conform more closely to that developed by the Wetland Classifi- 
cation Committee (Martin et al. 1953). 
56 
