• Diet, breeding success, detectability, and density of the great horned owl (Bubo virginianus) at its northern range limit

      McConnell, Madison H.; Kielland, Knut; Breed, Greg; Shook, John (2019-05)
      I studied the diet, breeding success, detectability, and density of great horned owls (Bubo virginianus) in the Middle Fork of the Koyukuk Valley in Arctic Alaska. The study extended from the southern slopes of the Brooks Range to latitudinal tree line, the northern breeding limit of the species, and included what are likely to be the northernmost great horned owl nests on record (up to 68.0113 degrees north). I completed the study during the 2017 and 2018 breeding seasons, during years of high snowshoe hare (Lepus americanus) abundance. The focus of this study was to gain an understanding of how high snowshoe hare abundance influences the recruitment, diet, and distribution of this apex generalist predator, and to determine best methods of detecting great horned owls for similar studies in the future. I used motion sensor cameras on nests as well as pellet analysis for diet and breeding studies, and call surveys for information on detectability and density. Great horned owl diet consisted mostly of snowshoe hares by mass (mean 80%, range 65-99%), with an average prey size of 714 g (95% CI ± 34.26). Nestlings received an average of 459 g (95% CI ± 75) of prey per chick per day, and the proportion of hares in their diet positively correlated with fledging success (P = 0.01). During call surveys, length of playback was the most important factor in detecting great horned owls throughout 12 minute surveys, reaching 23% (95% CI = ± 6.4) at 3 minutes, and up to 80% (95% CI = ± 6.1) at 9 minutes. Inclusion of silent listening periods may lessen the chance of detecting great horned owls during playback surveys, though a larger sample size is needed (P = 0.18). There was no correlation between cloud cover and probability of detection (P = 0.60) or wind speed and probability of detection (P = 0.28). However, there was a positive correlation between temperature and probability of detection (P = 0.02). Call surveys gave an estimate of 4.1 great horned owls per square kilometer (z = 4.302, 95% CI = ± 2.63). This was the northernmost study of North America's most widespread year-round bird of prey, and the first density estimate at their northern breeding limit.