Document Type

Thesis - Open Access

Award Date


Degree Name

Master of Science (MS)


Wildlife and Fisheries Science

First Advisor

Lester D. Flake

Second Advisor

Charles G. Scalet


Waterfowl broods on stock ponds in south central South Dakota were surveyed in 1977 and 1978 to compare flush and observation techniques for censusing broods and to determine waterfowl brood habitat preferences. Results indicated that 72-8% of the broods present were censused by either technique alone. No differences were indicated between the 2 techniques when analyzed by species. For total brood counts, a significant (P<0.05) difference was indicated for 1 of the 4 sample periods. Differences were found between the 2 techniques for certain age-classes of species and age-classes overall. Visibility correction factors were calculated for each species based on the percentage of known broods censused by each technique. Brood habitat preferences were examined using multiple regression and discriminant analyses of 33 pond and upland variables. Ponds were selected in a limited size range (0.71-2, 70 ha) to reduce variation in numbers of broods per pond due to pond size. However, variables indicative of pond size were significantly and positively associated to brood use of ponds in 8 of 9 analyses. Potential brood escape cover such as Polygonum spp. and Eleocharis spp. were positively associated with mallard (Anas platyrhynchos), (A. strepera), northern shoveler (A. clypeata) and class 1 broods. Large differences in numbers of broods on study ponds between the 2 years were likely due to increased nesting cover provided by yellow sweet clover (Melilotus officinalis). Study ponds with higher pH and more submersed aquatic vegetation had more brood use.

Library of Congress Subject Headings

Waterfowl -- South Dakota
Bird populations -- South Dakota
Farm ponds -- South Dakota


Includes bibliographical references (pages 38-42)



Number of Pages



South Dakota State University


No Copyright - Non-Commercial Use Only