Roost site characteristics of Northern spotted owls in the nonbreeding season in Central Washington

Roost site characteristics of Northern spotted owls in the nonbreeding season in Central Washington

Herter, Dale R

ABSTRACT. To evaluate habitat important to northern spotted owls (Strix occidentalis caurina) away from the nest stand, we used radio telemetry to locate adult owls at diurnal roosts during the nonbreeding season (September 1-March 15, 1994-1997). We recorded physiographic variables and measured within-stand structural characteristics within nested circular plots centered on roost trees. We then compared owl use plots to random plots selected within suitable habitat and inside the approximated home ranges of the 14 (7 male, 7 female) owls studied. Spotted owls selected older forest over young forest. Owls selected sites lower in elevation, with larger tree diameter at breast height (dbh), fewer trees/ha, greater canopy cover, less cover of low shrubs, and fewer pieces of down wood than random locations. Females used old-growth and mature forest to a greater degree than males. Fewer trees/ha, less cover of low shrubs, and fewer pieces of down wood/transect best discriminated roost sites from random sites. These characteristics are not indicative of stand conditions thought to maximize prey density. Results may indicate selection of specific sites for roosting. However, because spotted owls opportunistically take prey from diurnal roosts and often roost in foraging stands, they may also be selecting for site characteristics that facilitate the capture of prey. FOR. Sci. 48(2):437-444.

Key Words: Strix occidentalis caurina, habitat, logistic regression, radio telemetry, resource selection probability function.

WITHIN-STAND STRUCTURAL CHARACTERISTICS Of forest habitats used by northern spotted owls during the nonbreeding period are generally not well studied, are confounded by studies that lump breeding and nonbreeding data together (e.g., North et al. 1999, Irwin et al. 2000), or are assumed to be similar to those used during the breeding season. Structural characteristics thought to be important to owls during the breeding season have been evaluated in several parts of the subspecies’ range (Solis and Gutierrez 1990, Mills et al. 1993, North 1993, Buchanan et al. 1995, Thome et al. 1999). However, telemetry studies of owl pairs that maintain year-round territories indicate that pairs spend increasing amounts of time apart following the breeding season and typically expand their foraging area (Forsman et al. 1984, Carey and Peeler 1995). The nonbreeding season represents a time when habitat studies of roosting owls would not be influenced by the propensity for adult owls to return to the nest stand during daylight to guard the incubating bird or fledglings.

Northern spotted owls typically select mature and oldgrowth forest stands for nesting and brood-rearing (Forsman et al. 1984, Carey et al. 1992, Guti6rrez et al. 1995). In Douglas-fir (Pseudotsuga menziesii) and western hemlockdominated (Tsuga heterophylla) forests of the Pacific Northwest, northern spotted owls may use landscapes with extensive cover of young trees (

We used radio telemetry to study within-stand forest structural characteristics at roost sites of northern spotted owls during the nonbreeding season. We tested two hypotheses: (1) spotted owls select within-stand forest characteristics that optimize the abundance of small mammal prey because of their propensity to roost in foraging areas, and (2) differential habitat use by male and female spotted owls may continue or even increase during the nonbreeding period. Recent investigations have concluded that several within-stand features such as presence of snags, understory shrub cover, and amount of down wood are important to some small mammal prey [e.g., northern flying squirrels (Glaucomys sabrinus) and deer mice (Peromyscus spp.)] in Northwest forests (Carey 1995, Carey and Johnson 1995). Retention of snags and down wood during timber harvest is currently required by Washington State forest practices regulations (Washington Administrative Code 222-16). Moreover, studies of spotted owls (Solis and Gutierrez 1990) and other sexually dimorphic raptors have suggested that males and females may use habitats with different vegetative structure.

Understanding seasonal differences in roost site and foraging habitat selection by owls, as well as optimal habitat for their prey, will help managers to create or enhance stands for this species, which is federally listed as threatened. Private and public landowners concerned with timber production within the range of the spotted owl, as well as conservation of the species, may desire to contribute to the roosting and foraging needs of spotted owls by providing managed young forest stands useful to owls outside protected core nesting areas.

Study Area

This study was conducted within the western hemlock vegetation zone (Franklin and Dyrness 1988) on the western slopes of the Cascade Range in central Washington State. The owl territories we studied were mostly located in the upper Green and Greenwater River drainages, with one site at the headwaters of the Snoqualmie and Yakima Rivers along the Cascade Crest. Forest habitats at all study sites were similar; Douglas-fir/western hemlock forests predominated at low- to mid-elevations, intergrading with stands dominated by Pacific silver fir (Abies amabilis) and mountain hemlock (Tsuga mertensiana) at higher elevations. River and stream riparian zones contained some mixed and deciduous forests, which comprised

Methods

We attached radio-transmitters to 14 adult spotted owls, representing one or both members of a mated pair selected from 10 known, long-term (>5 yr) breeding territories. All 10 territories occurred in areas undergoing active forest management. Four of the territories overlapped one another

We recorded forest seral stages for each roost site and plotted locations on habitat maps. Habitat mapping was based on definitions provided by the State of Washington for suitable owl habitat (Appendix 4 in Hanson et al. 1993). We used this definition of habitat because it was the regulatory definition governing state and private timber harvest activities in spotted owl habitat in Washington. Habitat maps were produced from aerial photos (1:12,000) with a mapping resolution of approximately 0.5 ha and were field-verified. Two major seral stages were recognized as spotted owl habitat in the study area. Mature Forest habitat included oldgrowth and mature coniferous stands dominated by large (>50 cm dbh) overstory trees with >60% canopy closure. Young Forest habitat consisted of coniferous stands dominated by medium-sized trees (generally >20 cm dbh) with >60% canopy closure. All other areas comprised the NonHabitat category and included forest stands dominated by trees 70% canopy cover of deciduous trees (based on Hanson et al. 1993), stands above 1,372 m elevation, all sapling, seedling, shrub, forb, and grass dominated habitats, open water, and bare rock (Hanson et al. 1993). Old-growth was not separated as a category by itself because of its low abundance in the study area.

We recorded the elevation and aspect at each owl roost site from estimated locations plotted on topographic maps. To examine within-stand structural characteristics at roosts, we used a nested set of circular sample plots to obtain data on live trees, snags, and understory vegetation. The roost tree was located in the center of each plot. We used point and transect methods to measure tree canopy closure and down wood, respectively. During the first season of the study, we did not record percent cover of medium and tall shrubs or dbh of snags. We incorporated additional data on shrub layers and snag diameters in subsequent years for comparisons with other published studies (e.g., Mills et al. 1993, Carey and Johnson 1995).

We recorded the species and dbh for each live tree >10 cm encountered on a 0.04 ha circular plot. Overstory canopy cover (%) of the roost site was estimated using a moosehorn densiometer (Robinson 1947, Garrison 1949) at the four cardinal points 8 m from the roost tree. This distance was chosen in an effort to avoid undue influence from the roost tree itself. Because snags occurred at lower densities than live trees, we used a larger (0.08 ha) circular plot to record height and dbh of each snag >10 cm dbh. We used a scale of 1 (recently dead) to 5 (well decayed) to index the degree of decay for all snags measured (Cline et al. 1980). Percent of the ground covered by shrubs (including young conifers) was estimated within a 0.02 ha plot because of the difficulty in making ocular estimates on larger plots. Percent shrub cover was estimated for each of three layers: low (0-0.5 in), medium (0.5-2 in), and tall (>2m to the bottom of the tree canopy; adapted from O’Brien and Van Hooser 1983). To assess the amount of down wood present, we counted and estimated diameters of down logs (>3m min length and >10 cm in diameter) where they intersected a 20m in transect extending in a random compass direction from the plot center, and then calculated volume following Van Wagner (1968). We also used a scale of 1 (recently fallen) to 5 (well decayed) to index the degree of decay for each piece of down wood on the transect (Maser et al. 1979). Some sample size differences resulted from periods of heavy snow cover, which restricted our ability to obtain low shrub and down wood measurements.

To determine within-stand forest structural characteristics selected by owls, we compared measurements from owl diurnal roost sites with measurements taken at 60 random plots. Random plots were located using the intersections of a 1 km grid overlaid on a map of the study area, with each intersection numbered consecutively and chosen randomly. All random plot selections that fell in Non-Habitat polygons were discarded and another chosen until 60 plots within defined spotted owl habitat were selected. In the field, random plots were centered on the closest tree judged large enough to support a roosting owl. We restricted selection of random plots to suitable spotted owl habitat within a 2.9 km radius (2,636 ha) of the 10 owl territories, the area approximating the size of an annual pair home range for this region (Hanson et al. 1993).

Assuming our sample of roost sites was an unbiased sample, we compared the attributes of our roost sites to those from 60 randomly selected points (within owl habitat) in an exploratory analysis of roost-vegetation correlative associations. We used several statistical tests, both univariate and multivariate, to test use versus availability of forest seral stages and within-stand forest characteristics. We used a chi-square statistic to test the hypothesis that roost sites were distributed among mature and young forest habitats in proportion to their occurrence. We used one-way analysis of variance (ANOVA) to compare withinstand forest structural characteristics between the 60 random sites and 88 roost sites for which a complete set of forest structural measurements could be obtained. Selection of plots with all measurements may have biased our sample toward fall and early winter time periods when snow cover was light or nonexistent. Because of differences noted between male and female roost habitat, we also used ANOVA to compare 60 random sites and 40 male roost sites, 60 random sites and 48 female roost sites, and 40 male roost sites and 48 female roost sites. Variables that did not meet assumptions of normality and homoscedasticity were transformed using natural logarithms. In addition to measurements recorded in the field, we obtained an index of the heterogeneity of tree diameters by using the standard deviation of dbh measurements as an additional variable. To assess the volume of trees present near roost sites, we calculated quadratic mean diameter (QMD), a measure that estimates the diameter of the tree of average basal area (Husch et al. 1972). We transformed percent data (e.g., canopy closure) for all ANOVA analyses, using the arcsine of the square root to meet assumptions of normality. We used Rayleigh’s test (Batschelet 1981) to test for differences in angular distribution of (slope) aspect for owl sites versus random sites.

We collected habitat data at 175 owl roost sites. We excluded 29 roost sites from further analyses for the following reasons: (1) 10 fell outside the 2.9 km boundary from any site center (the area designated for random plot selection) and were used by one pair of owls that occasionally foraged up to 7 km from their site center; (2) 12 roost sites were found in areas mapped as nonhabitat (random plot selection was restricted to Mature and Young Forest habitat only); (3) and at 7 plots, both members of the pair were present, therefore we were unable to determine which sex had selected the roost site.

Habitat characteristics were analyzed using stepwise logistic regression to determine variables that best discriminated between owl and random sites. Because of differences found between male and female roost sites, we also used logistic regression to compare male roost sites with random sites, and female roost sites with random sites, as well as male versus female roost sites. Variables were selected for inclusion in the logistic regression analysis based on ANOVA results and pairwise comparisons. Variables with P values

Results

We summarized forest characteristics at 146 owl roost plots (73 male, 73 female) and 60 random plots in Table 1. Elevation and within-stand structural characteristics of live trees and understory shrubs differed between owl roost sites and random sites (Table 2). Within-stand structural characteristics of snags and down wood were not different. Owls used roost sites that were lower in elevation, with fewer trees/ ha, greater tree volume, greater variation in tree diameters, greater canopy cover, and less cover of low shrubs than random sites. Fewer trees/ha, less cover of low shrubs, and fewer pieces of down wood/transect best discriminated owl roost sites from random sites, based on logistic regression. The resource selection probability function for all owl sites versus random sites was calculated as: Pr^sub HAB^= 2.353 -0.004 (trees/ha) – 0.019 (low shrub cover) – 0.139 (pieces of down wood/transect).

Results of Rayleigh’s test on aspect data showed no differences (P

Comparisons between male or female roost sites and random sites showed that elevation, live tree characteristics, and some shrub measures were different. Canopy cover and low and medium shrub cover differed between male roost sites and random sites; male roost sites were lower in elevation with higher canopy cover and decreased low and medium shrub cover. For females, all five live tree characteristics, low shrub cover, and tall shrub cover influenced roost site use. Females used sites with fewer trees/ha, larger tree dbh, greater volume of trees, greater variation in tree diameters, and greater canopy cover, along with decreased cover of low shrubs and increased cover of tall shrubs. In logistic regression models, fewer trees/ha was the primary feature discriminating between male roost sites and random sites, with decreased medium shrub cover secondarily important. For females, greater variation in tree diameters, decreased cover of low shrubs, and increased cover of tall shrubs best discriminated female roost sites from random sites.

ANOVA models comparing sexes revealed that several of the roost site characteristics we measured differed significantly between males and females. All of the live tree variables differed between sexes except for canopy cover. Females selected roost sites with fewer trees/ha, greater volume of trees, larger tree dbh, and greater variation in tree diameters. Females also used sites with greater medium and tall shrub cover than males. Medium shrub cover was the most discriminating variable in comparisons between male and female roost sites (lower for males, higher for females) based on logistic regression analyses.

Discussion

Importance of Roost Habitat

Sites providing a safe location for diurnal roosting would be expected to play some role in the suite of habitats important to the spotted owl. Selection of a particular site or even a particular tree as a diurnal roost is likely an active decision on the part of this mostly nocturnal species. Owls may move from foraging sites to well-shaded sites in or below the canopy to meet thermoregulatory needs, avoid predation by larger diurnal raptors, and avoid harassment by corvids and other passerines (Barrows and Barrows 1978, Forsman et al. 1984). Selection of a site for roosting may not reflect the same habitat qualities necessary to meet foraging or reproductive needs. Because owls may often roost near sites used for foraging (Carey et al. 1989), habitats suitable for its major prey might also be expected to play a role in the overall suite of habitats important to this predator.

Flying squirrels and several smaller rodents comprise the bulk of prey for spotted owls in western Washington (E.D. Forsman, USDA Forest Service, personal communication). Recent habitat studies of sciurids (Carey 1995, Carey et al. 1999) and small mammals (Carey and Johnson 1995) in western Washington and Oregon have noted several within-stand structural characteristics that contribute to higher prey densities. Higher densities of large snags, increased ericaceous shrub cover (Carey 1995), and increased coarse woody debris (Carey et al. 1999) were important features correlated with greater use by northern flying squirrels. Small mammals were more common at sites with well-developed shrub communities and an abundance of coarse woody debris (Carey and Johnson 1995). We found that within-stand structural characteristics thought to be correlated with increased small mammal abundance were generally not important features at roost sites of spotted owls.

More large snags (snags/ha), increased cover of low shrubs (most of which were ericaceous shrubs), or increased cover of coarse woody debris were not important habitat components during the nonbreeding season. Instead, logistic regression models between owl plots and random plots indicated that fewer trees/ha, less cover of low shrubs, and fewer pieces of down wood/transect were the key habitat features discriminating owl use sites. Neither tree volume nor tree diameters contributed significantly, suggesting lower tree density was not strictly correlated with larger, older trees. Even in young stands, owls selected sites with more widely spaced trees. In other areas of western Washington, both Mills et al. (1993) and North (1993) noted tree canopy diversity and greater snag diameter were important for discriminating owl sites during the breeding season. We did not attempt to estimate tree canopy diversity and none of the snag characteristics we measured played an important role in defining roost sites. Our study area had a widespread and relatively high density of snags, probably due to extensive fires in the early 1900s (Plummer 1902); therefore, if owls selected sites for roosting based on proximity to snags, we were unable to detect it. Less cover of low shrubs at owl roosts might be explained by an owl’s preference for trees providing shady microsites within the forest. However, differences in mean canopy cover between owl roost sites (84%) and random sites (78%) were minor. Our findings suggest that roost sites used during the nonbreeding season may differ both from breeding season habitat, and habitat for some of the major prey species of spotted owls.

Relationship of Roost Sites to Foraging Habitat

Spotted owls are sit-and-wait predators, and when not actively nesting or raising young, often remain in a stand used for foraging for more than a day (Carey et al. 1989, Carey and Peeler 1995). Nocturnal triangulations on four of the owls we studied (two males, two females) during the first two seasons were regularly obtained. Roost sites of these owls were typically located in the same stand the owl had been using or would use during hours of darkness (and therefore presumably foraging). Owls are also known to capture natural prey (Sovern et al. 1994) or pen-raised prey (Forsman 1983) available to them at diurnal roosts. We observed roosting owls capturing natural prey during daylight on two occasions. In our study area, selection of a roost site may reflect choices in adjacent foraging habitat, particularly during the nonbreeding season when owls do not regularly return to roost near the nest tree.

Spotted owls capture prey both in trees and on the ground (Forsman et al. 1984), Selection of more widely spaced trees for ease of movement, and sites with unimpeded access to the forest floor (less down wood, fewer low shrubs) may have been reflected in our data at roost sites. Selection of habitat conducive to hunting rather than habitat with higher prey densities has been noted previously for spotted owls and other raptors. In northern California, dusky-footed woodrats (Neotoma fuscipes) are a major prey of spotted owls, and owls have been noted hunting in older forests along the edges of dense young stands (Zabel et al. 1995). Woodrats were more common in brushy sapling and pole stands (Sakai and Noon 1993), but owls preferred to capture them in adjacent old forest stands. In England, Southern and Lowe (1968) found tawny owls (Strix aluco) caught proportionally more woodmice (Apodemus sylvaticus) in areas of sparse shrub cover than in areas of dense shrub cover. Wood-mice were found at greater densities in dense shrub cover, but tawny owls were apparently less able to capture them in that cover type. Studies of northern spotted owls (Zabel et al. 1995) and northern goshawks (Accipiter gentilis; Widen 1989) concluded that some forest habitats were too dense for successful hunting whereas others were apparently too open, allowing the prey to detect the predator at an earlier point during capture attempts.

Sex Differences in Habitat Use

Male spotted owls roosted at sites where within-stand structural characteristics, and to some degree seral stage (reflected in X^sub 2^ values), differed from those used by females. Our results also indicated females favored sites that were older (greater tree volume, larger tree diameters, and fewer trees/ha) and of greater habitat complexity (greater variation in tree diameters, increased cover of medium and tall shrubs) than males. Although the logistic regression models indicated only cover of medium shrubs discriminated male plots from female plots, tree volume and variation in tree diameters were secondarily important features (logistic regression; P = 0.09 and 0.11, respectively). Males were found at sites with smaller trees, less cover of medium shrubs, and less variation in tree diameters compared to females.

These results compliment those of Solis and Guti6rrez (1990) for northern spotted owls in California, and other raptors with significant reversed sexual size dimorphism, in which the smaller males tend to use younger, denser forest habitats than the larger females (e.g., Widen 1989). Hypotheses concerning the evolutionary causes of reversed sexual size dimorphism continue to elicit much theoretical discussion (Andersson and Norberg 1981, Mueller 1986, Hedrick and Temeles 1989). Nonetheless, for northern spotted owls, differential use of forest structural characteristics based on sex has been suggested by at least two studies and should be considered when interpreting habitat use data for this subspecies.

Habitat Management Implications

Several studies of northern spotted owls have shown older forests to be favored by this subspecies, a result we confirmed during the nonbreeding period in the managed landscape of our study area. However, there was some use of younger forests, particularly those with characteristics in common with old forests [e.g., widely spaced trees, dense canopies, several shrub layers present (see also Carey and Peeler 1995)]. The characteristics of these young stands used by owls may be useful for forest managers seeking to provide useable habitat in areas adjacent to patches of mature forest where owls nest (see Hicks et al. 1999). Data gathered on habitat characteristics in the nonbreeding season may help refine those forest conditions known to be important for spotted owls, and aid in formulating management strategies for areas with similar habitat regimes.

We found 5-8% of roost sites occurred in areas mapped as Non-Habitat under State of Washington definitions. These were predominately in areas of high canopy cover of hardwoods (primarily red alder). During the nonbreeding season, tolerance of hardwoods may be higher than recognized by current owl habitat definitions (>70% of total canopy cover; Hanson et al. 1993). Mean elevation at owl roost sites was lower than at random plots, indicating our selection of random plots up to 1,372 rn in elevation (using State of Washington definitions) may have been too high in this portion of their range. In our study area, owls favored lower elevations and were rarely found above 1,300 m. We have not located spotted owl nests above 1,200 m in the study area, suggesting breeding season use may also be concentrated at lower elevations.

Habitat management for predatory animals is inevitably more complicated than that of herbivorous species. Managers must not only consider the basic life history requirements of the predator under study, but also the life history requirements of their prey and habitat conditions conducive to successful capture of that prey (Wakeley 1978, Preston 1990). Habitat patches yielding the greatest energy gain per unit of hunting effort may not contain the highest prey densities (Royama 1971). Forest management aimed solely at providing conditions preferred by rodent prey, without attention to roosting and foraging needs of northern spotted owls, might result in stands of lesser value to this federally listed species. Heterogeneous understories with patches of prime hiding cover (low shrubs and down wood) and den sites (down wood, snags, or deformed live trees) for flying squirrels and small mammals, interspersed with patches of suitable roosting and possibly foraging habitat for owls (low tree density, sufficient canopy cover, less cover of low shrubs and down wood), may better meet the spatial requirements of spotted owls. Thinning of young stands to promote understory development for small mammal prey of spotted owls, as suggested by Carey (1995), could be modified to provide shaded patches with open understories for roosting and hunting. Further research on prey habitat requirements, studies of additional prey species [e.g., woodrats, snowshoe hares (Lepus americanus), and pikas (Ochotona princeps)], as well as studies focused specifically on foraging owls, would yield more insight on the suite of habitats important for conservation of the northern spotted owl.

Literature Cited

ANDERSSON, M., AND A. NORBERG. 1981. Evolution of reversed sexual size dimorphism and role partitioning among predatory birds, with a size scaling of flight performance. Biol. J. Linnean Soc. 15:105-130.

BARROWS, C., AND K. BARROWS. 1978. Roost characteristics and behavioral thermoregulation in the spotted owl. West. Birds 9:1-8.

BATSCELET, E. 1981. Circular statistics in biology. Academic Press, New York. 371 p.

BINGHAM, B.B., AND B.R. NOON. 1997. Mitigation of habitat “take”: Application to habitat conservation planning. Conserv. Biol. 11:127-139.

BUCHANAN, J.B., L.L. IRWIN, AND E.L. McCUTCHEN. 1995. Within-stand nest site selection by spotted owls in the eastern Washington Cascades. J. Wildl. Manage. 59:301-310.

CAREY, A.B. 1995. Sciurids in Pacific Northwest managed and old-growth forests. Ecol. Applic. 5:648-661.

CAREY, A.B., AND M. L. JOHNSON. 1995. Small mammals in managed, naturally young, and old-growth forests. Ecol. Applic. 5:336-352.

CAREY, A.B., AND K.C. PEELER. 1995. Spotted owls: resource and space use in mosaic landscapes. J. Raptor Res. 29:223-239.

CAREY, A.B., S.P. HORTON, AND B.L. BISWELL. 1992. Northern spotted owls: Influence of prey base and landscape character. Ecol. Monogr. 62:223-250.

CAREY, A.B., S.P. HORTON, AND J.A. REID. 1989. Optimal sampling for radiotelemetry studies of spotted owl habitat and home range. USDA For. Serv. Res. Pap. PNW-RP-416.

CAREY, A.B., J. KERSCHNER, B. BISWELL, AND L. DOMINGUEZ DE TOLEDO. 1999. Ecological scale and forest development: Squirrels, dietary fungi, and vascular plants in managed and unmanaged forests. Wildl. Monogr. 142. 71 p.

CLINE, S.P., A.B. BERG, AND H.M. WIGHT. 1980. Snag characteristics and dynamics in Douglas-fir forests, western Oregon. J. Wildl. Manage. 44:773-786.

COSTANZA, M.C., AND A.A. ADFIFI. 1979. Comparison of stopping rules in forward stepwise discriminant analysis. J. Am. Stat. Assoc. 74:777-785.

FORSMAN, E.D. 1981. Molt of the spotted owl. Auk 98:735-742.

FORSMAN, E.D. 1983. Methods and materials for locating and studying spotted owls in Oregon. USDA For. Serv. Gen. Tech. Rep. PNW-GTR- 162. 8 p.

FORSMAN, E.D. 1988. A survey of spotted owls in young forests in the northern Coast Range of Oregon. Murrelet 69:65-68.

FORSMAN, E.D., E.C. MESLOW, AND M.J. STRUB. 1977. Spotted owl abundance in young versus old-growth forests. Oregon Wildl. Soc. Bull. 5:43-47.

FORSMAN, E.D., E.C. MESLOW, AND H.M. WIGHT. 1984. Distribution and biology of the spotted owl in Oregon. Wildl. Monogr. 87. 64 p.

FRANKLIN, J.F., AND C.T. Dress. 1988. Natural vegetation of Oregon and Washington. Oregon State Univ. Press, Corvallis, OR. 452 p.

GARRISON, G.A. 1949. Uses and modifications for “moosehorn” crown closure estimation. J. For. 47:733-735.

GUTITRREZ, R.J., A.B. FRANKLIN, AND W.S. LAHAYE. 1995. Spotted owl (Strix occidentalis). Number 179 in The birds of North America, A. Poole and F. Gill (eds.). The Academy of Natural Sciences, Philadelphia, and the American Ornithologists’ Union, Washington, DC. 26 p.

HANSON, E., L. HICKS, L. YOUNG, AND J. BUCHANAN. 1993. Spotted owl habitat in Washington; a report to the Washington Forest Practices Board. Washington Forest Practices Board Spotted Owl Advisory Group, Final Report. Olympia, WA. 116 p.

HEDRICK, A.V., AND E.J. TEMELES. 1989. The evolution of sexual dimorphism in animals: Hypotheses and tests. Trends Ecol. Evol. 4:136-138.

HICKS, L.L., H.C. STABINS, AND D.R. HERTER. 1999. Designing spotted owl habitat in a managed forest. J. For. 97:20-25.

HOSMER, D.W., AND S. LEMESHOW. 2000. Applied logistic regression. Wiley, New York. 373 p.

HusCH, B., C.I. MILLER, AND T.W. BEERs. 1972. Forest mensuration. Wiley, New York. 410 p.

IRWIN, L.L., D.F. ROCK, AND G.P. MILLER. 2000. Stand structures used by northern spotted owls in managed forests. J. Raptor Res. 34:175-186.

MASER, C., R.G. ANDERSON, AND K. CORMACK, JR. 1979. Dead and down woody material. P. 78-95 in Wildlife habitats in managed forests: The Blue Mountains of Oregon and Washington, Thomas, J.W. (tech. ed.). USDA Agric. Handb. 553.

MEYER, J.S., L.L. IRWIN, AND M.S. BOYCE. 1998. Influence of habitat abundance and fragmentation on northern spotted owls in western Oregon. Wildl. Monogr. 139. 51 p.

MILLs, L.S., R.J. FREDRICKSON, AND B.B. MOORHEAD. 1993. Characteristics of old-growth forests associated with northern spotted owls in Olympic National Park. J. Wildl. Manage. 57:315-321.

MUELLER, H.C. 1986. The evolution of reversed sexual dimorphism in owls: An empirical analysis of possible selective factors. Wilson Bull. 98:387-406.

NORTH, M.P. 1993. Stand structure and truffle abundance associated with northern spotted owl habitat. Ph.D. Diss., University of Washington, Seattle, WA. 113 p.

NORTH, M.P., JR FRANKLIN, A.B. CAREY, E.D. FORSMAN, AND T. HAMER. 1999. Forest stand structure of the northern spotted owl’s foraging habitat. For. Sci.45:520-527.

O’BRIEN, R., AND D.D. VAN HOOSER. 1983. Understory vegetation inventory: An efficient procedure. USDA For. Serv. Res. Pap. INT-323. 22 p.

PLUMMER, F.G. 1902. Forest conditions in the Cascade Range, Washington, between the Washington and Mount Rainier Forest Reserves. USDI Geol. Surv. Prof Pap. No. 6, Series H, Forestry 3. 26 p.

PRESTON, C.R. 1990. Distribution of raptor foraging in relation to prey biomass and habitat structure. Condor 92:107-112.

ROBINSON, M.W. 1947. An instrument to measure forest crown cover. For. Chron. 23:222-225.

ROSENBERG, D.K., AND R.G. ANTHONY. 1992. Characteristics of northern flying squirrel population in young second- and old-growth forests in western Oregon. Can. J. Zool. 70:161-166.

ROYAMA, T. 1971. A comparative study of models for predation and parasitism. Res. on Popul. Ecol., Suppl. 1. 65 p.

SAKA, H.F., AND BR. NOON. 1993. Dusky-footed woodrat abundance in different-aged forests in northwestern California. J. Wildl. Manage. 57:373-381.

SOLIS, D.M., JR., AND R.J. GUTTERREZ. 1990. Summer habitat ecology of northern spotted owls in northwestern California. Condor 92:739-748.

SOUTHERN, H.N., AND V.P.W. LOWE. 1968. The pattern of distribution of prey and predation in tawny owl territories. J. Anim. Ecol. 37:75-97.

SOVERN, S.G., E.D. FORSMAN, B.L. BISWELL, D.N. ROLPH, AND M. TAYLOR. 1994. Diurnal behavior of the spotted owl in Washington. Condor 96:200-202.

THomE, D.M., C.J. ZABEL, AND L.V. DILLER. 1999. Forest stand characteristics and reproduction of northern spotted owls in managed north-coastal California forests. J. Wildl. Manage. 63:44-59.

VAN WAGNER, C.E. 1968.The line intersect method in forest fuel sampling. For. Sci. 14:20-26.

WAKELEY, J.S. 1978. Factors affecting the use of hunting sites by ferruginous hawks. Condor 80:316-326.

WIDEN, P. 1989. The hunting habitats of goshawks Accipiter gentilis in boreal forests of central Sweden. Ibis 131:205-231.

ZABEL, C.J., K. McKELVEY, AND J.P. WARD, JR. 1995. Influence of primary prey on home-range size and habitat-use patterns of northern spotted owls (Strix occidentalis caurina). Can. J. Zool. 73:433-439.

Dale R. Herter is Wildlife Biologist, Raedeke Associates, Inc., 5711 N.E. 63rd Street, Seattle, WA 98115-Phone: (206) 525-8122; Fax: (206) 5262880; E-mail: dherter@nwlink.com. Lorin L. Hicks is Director, Fish and Wildlife Resources, Plum Creek Timber Co., 999 Third Avenue, Suite 2300, Seattle, WA 98102-E-mail: (hicks@plumcreek.com. Henning C. Stabins is Wildlife Biologist, Plum Creek Timber Co., P.O. Box 1990, Columbia Falls, MT 59912- E-mail: hstabins@plumcreek.com. Joshua J. Millspaugh is Professor, The School of Natural Resources, University of Missouri, 302 Anheuser-Busch Natural Resource Building, Columbia, MO 65211-E-mail: millspaughj@missouri.edu. Amy J. Stabins is Wildlife Biologist, Raedeke Associates, Inc., 5711 N.E. 63rd Street, Seattle, WA 98115. Larry D. Melampy is Wildlife Biologist, 306 S. Sampson Street, Ellensburg, WA 98926.

Acknowledgments: The authors thank J. Bottelli, K. Kraft, M. MacDonald, and D. Malkin for assistance with field studies. Habitat maps were produced by M. Baumgartner, R. Early, T. Hitzroth, and R. Marx. R. Burgess, L. Diller, E. Forsman, L. Irwin, and two anonymous reviewers greatly improved later versions of the manuscript.

Manuscript received November 29, 2000. Accepted September 19, 2001.

Copyright Society of American Foresters May 2002

Provided by ProQuest Information and Learning Company. All rights Reserved