Natural Resource Ecology and Management
Journal or Book Title
Western North American Naturalist
Since 1993, elk (Cervus canadensis nelsoni) abundance in the Black Hills of South Dakota has been estimated using a detection probability model previously developed in Idaho, though it is likely biased because of a failure to account for visibility biases under local conditions. To correct for this bias, we evaluated the current detection probability across the Black Hills during January and February 2009–2011 using radio-collared elk. We used logistic regression to evaluate topographic features, habitat characteristics, and group characteristics relative to their influence on detection probability of elk. Elk detection probability increased with less vegetation cover (%), increased group size, and more snow cover (%); overall detection probability was 0.60 (95% CI 0.52–0.68), with 91 of 152 elk groups detected. Predictive capability of the selected model was excellent (ROC = 0.807), and prediction accuracy ranged from 70.2% to 73.7%. Cross-validation of the selected model with other population estimation methods resulted in comparable estimates. Future applications of our model should be applied cautiously if characteristics of the area (e.g., vegetation cover >50%, snow cover >90%, group sizes >16 elk) differ notably from the range of variability in these factors under which the model was developed.
Works produced by employees of the U.S. Government as part of their official duties are not copyrighted within the U.S. The content of this document is not copyrighted.
Phillips, Evan C.; Lehman, Chadwick P.; Klaver, Robert W.; Jarding, Angela R.; Rupp, Susan P.; Jenks, Jonathan A.; and Jacques, Christopher N., "Evaluation of an elk detection probability model in the Black Hills, South Dakota" (2020). Natural Resource Ecology and Management Publications. 341.