(1983 and 1985). Patchy snow cover (1982 and 1986) may result 

 in more variable observability. The Lincoln and modeled 

 population estimates were very close for early winter (Table 

 2.3), deviating by an average of 3.7%. 



Spring surveys were conducted with a helicopter during 

 1978-1980 and a Piper SuperCub from 1981 through 1987 (Table 

 2.5). The SuperCub surveys were generally flown a week or two 

 later than the helicopter surveys, shortly after "green-up" 

 when most deer concentrated on open ridgetops to feed on new 

 growth. Observability from the helicopter was high and similar 

 to early winter except during 1978 when survey conditions were 

 poor, the marked sample was low, and deer were actively moving 

 to their normal ranges from areas along the river where they 

 had concentrated during the severe winter. 



Average observability from the SuperCub was also high 

 (68.3%), and the coefficient of variation between years was 

 low (8.0%). Both cloud cover and stage of vegetation 

 "green-up" may influence observability from a fixed-wing 

 aircraft during spring. For example, slightly more than 60% 

 of the marked deer were observed in 1983 and 1986 when skies 

 were generally overcast, and the 1986 survey was conducted 

 before "green-up" was sufficiently advanced for maximum deer 

 use of open areas. Excluding those years, observability 

 averaged 71%. The average deviation between Lincoln and 

 modeled population estimates for spring was 6.5% (Table 2.3). 



The accuracy of our counts and percentages of marked deer 

 observed during winter and spring were at least as high and 

 usually higher than reported by others for mule deer (Bartmann 

 et al. 1986), white-tailed deer (Rice and Harder 1977, Floyd 

 et al.1979, Beasom et al . 1986), and moose (LeResche and 

 Rausch 1974). Variation in percentages of marked deer 

 observed during winter and spring among years was much less 

 than reported elsewhere for deer (Bartmann et al . 1986, Beasom 

 et al. 1986), but similar to that for aerial surveys of moose 

 conducted by experienced observers under excellent conditions 

 (LeResche and Rausch 1974). 



The high precision of observability indices for winter 

 and spring provided confidence in the accuracy of annual 

 population estimates for those seasons. For example, during 

 6 of the 7 winter surveys flown with pilot B, the proportion 

 of marked deer observed varied only from 0.77 to 0.83 (Table 

 2.4). During the seventh year, when survey conditions were 

 poor, observability was 0.68. If we assume a total count of 

 1,000 deer and apply a mean observability index of 0.80 (52 of 

 65 marked deer seen), our Lincoln estimate would be 1,250 

 deer. Considering the possible range in observability 

 recorded for most years, we might expect the total population 

 to lie between 1,200 (1,000/0.833) and 1,300 (1,000/0.769). 



37 



