Habitat Selection by Mexican Spotted Owls in Northern Arizona

Abstract

We compared use of seven habitat types to availability of those types within the home ranges of eight radio-tagged Mexican Spotted Owls (Strix occidentalis lucida). When all habitat types were considered simultaneously, habitat use differed from habitat availability for each owl. Patterns of habitat use varied among individuals and with respect to activity. Owls generally foraged more than or as frequently as expected in virgin mixed-conifer and ponderosa pine (Pinus ponderosa) forests, and less than expected in managed forests. Owls roosted primarily in virgin mixed-conifer forests. We also compared habitat characteristics among foraging, roosting, and randomly available sites. Habitat characteristics differed significantly among plot types. Both roosting and foraging sites had more big logs, higher canopy closure, and greater densities and basal areas of both trees and snags than random sites. Roosting sites had greater canopy closure, more big logs, and greater densities of both trees and snags than foraging sites. Mature forests appear to be important to owls in this region, and different forest types may be used for different activities. THE MEXICAN SPOTTED OWL (Strix occidentalis lucida) occurs throughout forested highlands and rocky canyonlands in the southwestern United States and Mexico (McDonald et al. 1991). This owl is often associated with virgin forests in northern Arizona (Ganey and Balda 1989a), but little is known about its habitat requirements or how timber harvesting might affect owl habitat. Populations of the closely related Northern Spotted Owl (S. o. caurina), which appear to be closely tied to old-growth coniferous forests in the Pacific Northwest (Forsman et al. 1984, Ca-rey et al. 1990, 1992, Solis and Guti6rrez 1990, Thomas et al. 1990, Blakesley et al. 1992), are declining as timber harvesting reduces the amount of such forest (Forsman et al. 1984, 1988). As a result, concern has arisen over the potential effects of timber harvesting on Mexican Spotted Owls and their habitat. Because of this concern, the owl was recently listed as Threatened (Tur-ner 1993). Information on specific patterns of habitat use by Mexican Spotted Owls is needed to evaluate the potential effects of timber harvesting on this owl. Here we describe patterns of habitat use within the home ranges of radio-tagged Mexican Spotted Owls (third-order selection; John-son 1980) at two different spatial scales. Specif-• Present address: USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, 2500 S. Pine Knoll, Flagstaff, Arizona 86001, USA. ically, we compare owl use of habitat types to the availability of those habitat types within owl home ranges, and identify habitat features consistently occurring in areas used by Spotted Owls. METHODS We monitored radio-tagged owls on three study areas in northern Arizona. The San Francisco Peaks (SFP) study area is located 3 km north of Flagstaff, the Walnut Canyon (WC) study area 4 km southeast of Flagstaff, and the White Mountains (WM) study area approximately 27 km southwest of Alpine in east-central Arizona. Elevations range from approximately 1,830 to 2,160 m at WC, 2,130 to 2,650 m at WM, and 2,190 to 2,930 m at SFP. All three areas have relatively cool summers with frequent rainfall, and cold winters with extended periods of snow cover. Vegetation on all three study areas is predominantly coniferous forest. Ponderosa pine (Pinus ponderosa) forest dominates at lower elevations and on south-facing slopes, with mixed conifer forest prevalent on north-facing slopes and at higher elevations. Mixed-conifer forest is dominated by Douglas-fir (Pseudotsuga menziesii) and/or white fir (Abies concolor). Eight adult owls were captured and radiotagged: both members of two pairs at SFP; both members of one pair at WC; and two males from adjacent drain-ages at WM. Methods and equipment used to capture, radio, and track Spotted Owls, as well as tracking periods for individual owls, were described in Ganey and Balda (1989b). Each time an owl was radio-located , we recorded the date, time, and activity type. We defined two activity types, assuming all day locations 162