TY - JOUR
AU - Bisset, Alan R.
AU - Rempel, Robert S.
PY - 1991/01/01
Y2 - 2022/01/24
TI - LINEAR ANALYSIS OF FACTORS AFFECTING THE ACCURACY OF MOOSE AERIAL INVENTORIES
JF - Alces: A Journal Devoted to the Biology and Management of Moose
JA - Alces
VL - 27
IS -
SE - Articles
DO -
UR - https://alcesjournal.org/index.php/alces/article/view/1111
SP - 127-139
AB - <p>Ontario has been using a standard aerial survey methodology since 1975 to estimate moose (<em>Alces alces</em>) abundance. Data have been collected on 13838 survey plots. Collection of flight and environmental conditions, in addition to the number of moose observed permits assessment of the importance of these factors to moose surveys. Calculation of multiple regression models included variables representing biological factors as well as sightability factors. Resurvey of 104 randomly selected quadrats immediately after the initial survey allowed more accurate assessment of sightability bias.</p><p>To explore the relationships between recorded variables and density of moose seen, multiple regression analysis was conducted on standard survey data. This analysis explained about 30% of the total variance in numbers of moose seen among plots. Variables were interpreted as those explaining biological and habitat factors, environmental factors, and flight condition factors. Many of the 19 variables contributed significantly to the models, although partial correlation coefficients were generally weak (r<sup>2 </sup>ranging from 0.0002 to 0.1747). Two variables interpreted as biological and habitat factors explained a total of about 8% of the variance. The most meaningful of the sightability variables was time-on-plot, accounting for 15% of the variability, with other sightability factors explaining less than 2% of the variability. Crew size and individual crew experience were important factors affecting the sightability of moose.</p><p>To calculate a linear correction coefficient, regression analysis was conducted on resurvey data from northwestern Ontario. Average sightability was 79% as estimated by the linear correction model. In this model only 5 variables were significant in accounting for 13% of the variance in survey results. The significant sightability factors were snow depth, crust, time spent on-plot during the initial survey, aircraft type, and cloud cover.</p>
ER -