25.09 per cent and fruits 11.79 per cent. The fruits were mostly 

 rose hips, but included hazelnuts and the acorns of scarlet oak. 

 Hamerstrom, Hopkins and Rinzel (1941:185) list cultivated grains, 

 weed seeds and the browse of trees as the important foods in winter 

 in Wisconsin. Schwartz (1945:72) found little evidence of the use 

 of browse of trees in Missouri, but com and sorghum were im- 

 portant winter foods. Yeatter (1943:414) reported ninety-one per 



cent plant food and nine per cent insects in summer in Illinois. 

 No reports, relating restrictedly to the Midwest, have been found 



listing foods used by prairie chickens prior to the growing of cul- 

 tivated crops there, but the early writings of Audubon (1834:491- 

 501) and Koch (1836:159) in reference to the prairie chicken in 

 Kentucky and Missouri, respectively, list fruit, tree buds, grain, 

 wild grapes, mistletoe, buds of willow and hazelnut, seeds of grass 

 and flowers (presumably forbs generally), and the catkins and 

 shoots of hazelnut. The German naturalist, Koch (loc. cit.), de- 

 scribed conditions in the state of Missouri soon after it was settled, 

 and mentions that the "cupido-huhn" moved from the prairies to 

 the cultivated land along the Missouri River in autumn. 



In the present study, information concerning foods and feeding 

 habits of the greater prairie chicken was obtained by observing the 

 birds in their feeding activities, by examining crop- and gizzard- 

 contents and by studying droppings found in the Welda Area. 



Crop- and gizzard-contents were washed, dried and analyzed 

 by separating the food items and measuring the volume of each 

 kind of food. Samples made up mostly of insects were preserved 

 in alcohol and measured in a moist condition. Droppings were 

 collected as entire deposits from a roost, and as such would contain 

 remains of most, or all, of the foods eaten late in the previous day. 

 Only fresh droppings were collected in order to make certain of the 

 date of deposition. Droppings were washed free of urates and 

 foreign material and then dried. The remains of the various food 

 items were identified and an estimate was made of the percentage 

 of each. The total volume of each sample was measured and the 

 volume of each item was computed from the estimated percentage. 

 This procedure is less exact than the analysis of crop- and gizzard- 

 contents, but does yield a broad picture of the trends of feeding 

 habits throughout the year that cannot be obtained otherwise, with- 

 out sacrificing many birds. Sixty-five samples of droppings, 29 

 crops and 20 gizzards were analyzed. 



In the droppings the remains of the more common seeds could 

 be classified and insect material usually could be classified to 

 family. Dicotyledonous leaves usually retained their shape, and 



[44] 



