Assessing aerial survey methods to estimate elk populations: a case study

Balancing the need for accurate estimates of wild ungulate populations with survey costs continues to challenge wildlife biologists. We present a case study to evaluate the feasibility of using stratified sampling and sightab ility models to estimate the population size and composition of elk (Cervus elaphus) in northeast Oregon in 1997. We conducted 4 helicopter sightabili ty surveys (3 winter and one spring) and one fixed-wing survey in spring. W e estimated population sizes +/- 90% confidence intervals as 1,964+/-310, 1 ,825+/-133, and 1,768+/-334 For the winter surveys, and 2,275+/-383 for the spring survey. Population estimates did not differ (P=0.28) among the 4 he licopter surveys. We also applied a sightability correction model to the ob served numbers on the fixed-wing survey and estimated a spring population o f 2,071+/-155. Raw counts of fixed-wing (n= 1,747) and helicopter (n=1,743) surveys in spring were nearly identical, but due to the sightability corre ction, results from the helicopter survey were adjusted upward by 31%, wher eas results from the fixed-wing survey increased 19%. Observed calf:cow rat ios during spring were similar between the fixed-wing survey (27.7:100) and the helicopter survey (27.3:100). Confidence intervals associated with hel icopter sightability model estimates of bull:cow ratios ranged from 33% to 85%. Properly designed fixed-wing surveys flown under appropriate condition s may provide population and calf:cow ratio estimates equivalent to helicop ter surveys at significant cost savings.