Habitat Use by Saw-Whet Owls in the Sierra Nevada

Authors: J. E. Groce, M. L. Morrison

Northern saw-whet owls (Aegolius acadicus) are secondary cavity-nesters and their dependence on snags has potential repercussions on forest management practices. Descriptive studies exist regarding habitat characteristics around saw-whet nest and roost areas, yet few studies have examined associations within larger areas or relative to snag characteristics (e.g., density). We conducted owl broadcast surveys and snag sampling during the spring and summer of 2006 and 2007 in the Lake Tahoe Basin of the central Sierra Nevada; we measured additional habitat variables from Geographic Information System layers. We modeled detection and occupancy probabilities for saw-whets using sampling and site covariates at survey sites. In addition, we used stepwise logistic regression to compare habitat characteristics at owl use sites and nonuse sites at 2 spatial scales. Detection probability was low in 2006 and decreased throughout the survey period; detection probability was slightly higher in 2007 and unaffected by day of survey. Probability of occupancy was affected by elevation and dominant tree species in 2007. Similarly, stepwise logistic regression indicated saw-whet occurrence was negatively correlated with the percentage of area dominated by white fir (Abies concolor) at both the macrohabitat (approx. 260 ha) and microhabitat (approx. 20 ha) scales and was positively correlated with the percentage of area containing open canopy at the microhabitat scale. We did not find correlations between saw-whet occurrence and snag characteristics. Current restoration projects in areas of the Sierra Nevada aim to decrease relative abundance of white fir and the number of snags in forest stands. We recommend continued monitoring of saw-whets to understand potential effects of these restoration activities. Our estimates of saw-whet occupancy and detection probabilities can be used by forest managers to determine necessary survey effort for reliable results when developing monitoring protocols.

Suggested Citation