example, if eighty percent of a thousand observations of species A were made in 

 habitat B, we can make such interpretations as: eighty percent of species A i 

 occurs in habitat B; species A uses habitat B eighty percent of daylight hours, 

 even though the actual number of species A present was two hundred and only one 

 hundred of these were actually observed. Such extrapolations and interpretations, 

 however, must be viewed with caution, with full knowledge of the sample bias 

 noted above, especially in cases where sample size is small. 



As the study progressed, attention was focused upon certain species or 

 groups of species of special concern, and a special effort was made to locate 

 and describe critical sites which are used from year to year such as leks, nest 

 sites, winter concentration areas, rookeries, and prairie dog towns. Owl pellets, 

 red fox and coyote scats, and crops of road-killed or hunter-killed gallinaceous 

 birds were collected when encountered and stored by DNRC, but no special effort 

 was made to determine food habits, as the literature provides adequate infor- 

 mation. Standard study skins were prepared of all bird and mammal specimens ob- 

 tained during the study; amphibians were preserved in a five percent formalin 

 solution. Data on all bird nests located during the study were recorded on 

 standard next-record cards (Appendix D) prepared by the Cornell Laboratory of 

 Ornithology. 



ROADSIDE WILDLIFE SURVEY AND ANALYSIS 



In addition to the general ground surveys, which were certainly not system- 

 atic and which do not provide information on relative or absolute abundances, 

 a method was sought which would provide quantitive information as well as sample . 

 a wide range of habitats without consuming great amounts of field time. The ' 

 standard nationwide Breeding Bird Survey (BBS) was chosen as the ideal method to 

 meet these goals. This method has been described and critiqued by Peterson 

 (1975), Robbins and Van Velzen (1967, 1969), Van Velzen and Robbins (1971), and 

 Weber and Theberge (1977); Lewis et al. (1978) and Swenson (1978) have also ap- 

 plied the method to baseline study. The BBS has been underway nationwide 

 since 1968, and in 1977 over 1700 such surveys were run in 48 states. The 

 opportunity is thus provided to tie in to a vast amount of data obtained by 

 standardized methods. While the technique was originally designed to sample 

 breeding season bird populations, it was easily applied as a roadside wildlife 

 survey without modification for all seasons of the year. 



Each roadside wildlife survey route consisted of 50 roadside stopping 

 points separated by 0.80 km (0.5 mi) as indicated by odometer readings, for a 

 total route length of 39.4 km (24.5 mi). Beginning 0.5 hr before local sun- 

 rise at the first stopping point of each route, the observer stood quietly 

 outside the vehicle and recorded the species and numbers of all vertebrates 

 seen or heard during a 3.0 minute counting period. All reptiles, am- 

 phibians, and mammals were recorded regardless of distance; birds were only 

 counted within an estimated 0.40 km (0.25 mi) radius. During winter and early 

 spring, most locations were made visually and the observer often walked as far 

 as 30 m (100 ft) from the vehicle while scanning the area for animals. Since 

 very few observations were made at each stop during this period, no difficulty 

 was encountered by the observer in timing stops or recording observations while 

 scanning. During the breeding season, however, most observations were based on 



28 



