More info for the terms: association, basal area, cover, density, frequency, fuel, hardwood, layering, litter, mesic, presence, selection, shrub, shrubs, snag, tree, xeric
According to reviews, spotted owls occur in closed-canopy, uneven-aged, late-successional and old-growth forests [84,85,193]; Mexican spotted owls also occur in deep, steep-walled canyons with little canopy cover [20]. Many habitat measurements were taken in plots between 0.1 and 2 acres (0.04-0.8 ha). In this section, these will be referred to as "small plots."
Elevation: Spotted owls occur at a range of elevations, with higher elevations occupied at lower latitudes.
Northern spotted owls occur at elevations from 70 to 6,600 feet (20-2,010 m), with the majority in the lower portions of this range [32,62,84]. In coniferous forests of northwestern California, nest sites ranged from 118 to 4,944 feet (35-1,507 m), with 94% occurring below 4,000 feet (1,218 m) [114]. In mixed evergreen and mixed-conifer forests of northwestern California, roosting northern spotted owls avoided areas above 2,950 feet (900 m) [23]. In coniferous forests of the Klamath, Coast and Cascade regions in Oregon and the Olympic peninsula of Washington, nest locations were significantly lower (P<0.001) in elevation than random sites within northern spotted owl's home ranges [91]. In coniferous forests of southwestern Washington, important owl locations (e.g., nest sites, multiple detection sites) averaged 3,170 feet (966.2 m), which was significantly (P<0.001) lower than the 3,510-foot (1,070.3 m) average elevation at random sites [161]. In coniferous forests of the eastern Cascade Range of Washington, elevation of northern spotted owl nest sites was negatively associated with latitude (P<0.001) [32], and site occupancy and reproductive rates were inversely associated with elevation [104].
According to reviews, California spotted owls occur on sites from about 1,000 to 8,500 feet (300-2,600 m), with individuals in southern California generally occurring at higher elevations [84,197]. A detailed summary of California spotted owl habitat associations reports nest sites occurring at an average elevation of 5,300 feet (1,620 m) in the northern Sierra Nevada and 6,000 feet (1,830 m) in southern California [85]. In white fir-mixed-conifer stands of the Lassen National Forest in northeastern California, elevation at California spotted owl nest areas was inversely associated with site occupancy and reproductive output [24].
According to a review and analysis of habitat relationships in the Mexican spotted owl recovery plan, 95% of Mexican spotted owl nests in Arizona and New Mexico study sites were in trees at elevations from 6,000 to 8,500 feet (1,829-2,591 m), with 72% occurring from 6,500 to 7,500 feet (1,982 and 2,287 m) [78]. In Colorado, Mexican spotted owls occurred from 5,820 to 9,100 feet (1,770-2,770 m), with 17 of 20 from 6,500 to 7,800 feet (1,980-2,380 m) [49]. In Saguaro National Park, Mexican spotted owl territories did not occur below 7,000 feet (2,100 m) [205]. In another southern Arizona study site, Mexican spotted owl nest/roost sites occurred from 5,820 to 7,620 feet (1,773-2,323 m) [46].
Topography: Effects of aspect on northern spotted owl habitat selection are ambiguous. Aspects at nesting and roost sites in mixed evergreen and mixed-conifer forests of northwestern California did not differ from availability [23]. In the western Cascade Range of southwestern Washington, aspects of spotted owl sites did not differ significantly (P>0.05) from random sites [161]. Similarly, data from low to mid-elevation forests of northwestern California also indicated that aspects at nest sites did not differ from random sites [114]. In southwestern Oregon, north to east aspects were used more frequently and south to southwest aspects were used less frequently than expected in summer. In contrast, spring and fall roost sites occurred more frequently on south and southwest aspects (P<0.05) [42]. The aspect at nest sites in the eastern Cascade Range of Washington averaged 35° northeast, which was significantly (P=0.015) different from the average aspect of 48° northeast on random sites in the nest stand [32]. Over 50% of summer roosting and foraging observations occurred on north-facing slopes in predominantly mixed-evergreen forests of northwestern California [180].
California and Mexican spotted owls may select northern slopes and/or avoid southern slopes. In the central Sierra Nevada, roost sites, but not nest sites, faced north (x=16° north) significantly (P<0.05) more often than a uniform distribution [22]. One of the many factors associated with higher reproductive rates in oak woodland of the southern Sierra Nevada was nesting on north-facing slopes [157]. Although likely confounded with vegetation characteristics, California spotted owl nests did not occur on sites with southern aspects in foothill riparian and oak woodlands in the southern Sierra Nevada [184]. However, the average aspects of nesting and roosting sites in the San Bernardino Mountains were not significantly different than average aspects of random sites, according to a detailed summary of California spotted owl habitat studies [85]. In southern Arizona, 7 out of 10 of Mexican spotted owl nest/roost sites were on northwest facing slopes [46]. In Saguaro National Park most roost sites occurred on northwest facing slopes [205]. According to a review and analysis of habitat relationships in the Mexican spotted owl recovery plan, nearly 50% of Mexican spotted owl nests occurred on north or northeast aspects in Arizona and New Mexico study sites [78].
Spotted owls often occur on steep slopes, and sometimes steep slopes are selected more than would be expected based on their availability. For example, in coniferous forests of the western Cascade Range in Washington, slopes at northern spotted owl sites averaged 54.1%, significantly (P<0.001) steeper than the 46.2% average slope on random sites [161]. In mixed evergreen and mixed-conifer forests of northwestern California, gentle (15-30%) slopes were avoided (P<0.03) for roosting [23]. Slope averaged 49% at roost sites in southwest Oregon [42] and 58% at nest sites in low- to mid-elevation coniferous forests of northwestern California [114], but these slopes were typical of the area. In mixed-evergreen and mixed-conifer forests of northwestern California, steep slopes were used by nesting northern spotted owls in proportion with availability [23]. According to a review, California spotted owl nesting and roosting sites were significantly (P<0.001) steeper than random sites in the San Bernardino Mountains [85]. Slopes of nest sites in foothill riparian and oak woodlands of the southern Sierra
Mexican spotted owl.
US Fish and Wildlife Service
Nevada ranged from 0% to 105% [184]. Mexican spotted owl nest areas in Arizona were significantly (P<0.001) steeper (x=38.5%) than random sites (x=20.6%) [70]. In steep, rocky, canyons of southern Arizona, the average slope of nest and roost sites was 34% [46].
Spotted owls seem to select the lower portions of slopes, at least in summer. Although most of the data are from studies on the northern spotted owl [23,42,59,62,91,114], there is evidence that California [184] and Mexican spotted owls [78,93] also select slope bases. In mixed-evergreen and mixed-conifer forests of northwestern California, northern spotted owls nested and roosted on the lower third of slopes significantly more, the middle third in proportion with, and the upper third of slopes significantly less than expected based on availability during the breeding season (α=0.05) [23]. On sites in coniferous forests in Oregon and Washington, 95% of nest sites were on the bottom or middle third of slopes, although this was only significantly (P<0.025) more than random sites within home ranges in the Klamath Mountains [91]. In southwestern Oregon, lower slopes were used significantly more than expected for roosting during summer. However, in the spring, fall, and winter northern spotted owl used upper and mid-slopes significantly more than expected (P<0.01) [42]. In foothill riparian and oak woodlands in the southern Sierra Nevada, California spotted owl nest sites were typically on the lower third of slopes [184]. Foraging Mexican spotted owls in mixed-conifer, mixed-conifer/oak, ponderosa pine, and pinyon-juniper habitats of southwestern New Mexico selected canyon bottoms or the lower portion of slopes significantly (P<0.05) more than expected in 3 study areas [93]. According to a review of habitat relationships in the Mexican spotted owl recovery plan, nearly half of Mexican spotted owl nests on Arizona and New Mexico study sites occurred on the lower third of slopes [78].
In coniferous forests of the eastern Cascade Range of Washington, topography may influence northern spotted owl reproductive success. In glacially-scoured, topographically-varied landscapes, average reproductive output was 0.38 fledglings/year, while in more gently rolling, montane slopes it averaged 0.57 fledglings/year (P<0.01) [104].
Water: In some regions, northern spotted owls use areas near water. In mixed-evergreen forests of northwestern California, the summer roost sites of 10 northern spotted owls averaged 466 feet (142.1 m) from water, which was significantly (P<0.01) shorter than the average 743 feet (226.6 m) from random locations to water [180]. In managed timberlands in the coastal redwood vegetation zone of northwestern California, northern spotted owl nest areas were closer to water than randomly-selected plots (P=0.032) [59]. Nest sites in low- to mid-elevation conifer forests of northwestern California averaged 385 feet (117.3 m) from water [114]. On 2 sites in the Coast and Cascade Ranges in western Oregon, 84% of nests were within 820 feet (250 m) of a stream or spring [62]. In southwestern Oregon, roost sites were significantly (P<0.01) closer to water in summer (x = 240 feet (74 m)) than in winter (x = 325 feet (99 m)) [42]. A literature review states that Mexican spotted owls occur in canyons with perennial water sources [83].
Weather: Reproductive rates of northern [64,104] and California [157] spotted owls are strongly influenced by weather. Low northern spotted owl reproductive output in Douglas-fir/hardwood and mixed-conifer forests of northwestern California was associated with cold, wet springs [64]. Average productivity in the eastern Cascade Range of Washington also declined with increasing precipitation, from 0.10 young/year on sites receiving more than 118 inches (300 cm) of precipitation a year to 0.96 young/year on sites receiving less than 20 inches (51 cm) of precipitation a year [104]. High California spotted owl reproductive rates were associated with less precipitation and higher minimum temperatures during the breeding season (March-May). For example, in mixed-conifer forests, reproductive output averaged 1.585 fledglings/pair in breeding seasons with less than 8 inches (20.7 cm) of rain and 0.307 fledglings/pair in breeding seasons with more than 8 inches of rain. In breeding seasons with more than 8 inches of rain, California spotted owl reproductive output was greater when the minimum April temperature was above 28 °F (-2 °C) (0.473 fledglings/pair) than when it was below 28 °F (0.183 fledglings/pair). The number of pairs breeding in each of these situations was small, ranging from 2 to 7 [157]. The similarity in California spotted owl reproductive trends over large portions of the southern Sierra Nevada also suggests weather is an influential factor [185]. There are no data on the effects of weather on Mexican spotted owl reproduction.
Survival and occupancy rates of northern spotted owls may also be affected by weather. Northern spotted owl survival in Douglas-fir/hardwood and mixed-conifer forests of northwestern California was detrimentally affected by cold, wet springs [64]. In coniferous forests of the eastern Cascade Range of Washington, precipitation was inversely related to site occupancy [104]. There is no information on the effect of weather on California or Mexican spotted owl survival or site occupancy.
Good quality habitat likely buffers the effects of weather. For example, data from Douglas-fir/hardwood and mixed-conifer forests of northwestern California suggest that decreases in northern spotted owl survival associated with cold, wet weather were more gradual in landscapes with features considered high quality compared to landscapes with low-quality features [64]. The habitat features associated with nest sites and the negative impact of precipitation on California spotted owl reproduction led researchers to speculate that high canopy cover and foliage volumes could reduce throughfall precipitation and wind penetration at nest sites [157].
Weather may also influence spotted owl's habitat selection. In Saguaro National Park, the average daytime temperature at Mexican spotted owl roosts was significantly cooler than the surrounding ambient temperature (P<0.05), with an average difference between roost and ambient temperatures of 5.1 °F (2.9 °C) [205]. In northern Arizona, nesting sites had significantly (P<0.001) lower temperatures and were above 95 °F (>35.2 °C) less often than randomly-selected areas. Cooler sites were associated with increased canopy cover (P=0.001) [70]. According to a review, canyons occupied by Mexican spotted owls in Zion National Park had higher humidity than canyons where owls were not detected, and roost sites in Utah occurred in canyons with lower temperatures than randomly-selected canyons [78]. North and others [157] suggest that the characteristically high foliage volume at California spotted owl nest sites in oak woodland and mixed-conifer forests of the southern Sierra Nevada may reflect selection for microhabitats that provide the most cover in inclement weather.
Northern spotted owls in western Oregon roosted higher in the canopy in cold wet weather than in warm or hot weather [62]. However, relationships between roost sites and climatic variables were weak in mixed-conifer and Douglas-fir forests in southwestern Oregon [42]. Summaries of the role of northern spotted owl thermoregulation in the selection of old-growth habitat are included in [41,83].
Succession: Northern spotted owls are strongly associated with mature and old-growth forests [23,42,62,100,105,161,164,180]. In Douglas-fir/hardwood, mixed-conifer, and Oregon white oak forests of California, 500-acre (200 ha) plots centered on nesting and roosting sites contained significantly (P=0.003) more mature and old-growth habitat than random plots [100]. In mixed-evergreen forests of northwestern California, northern spotted owls preferentially selected (P≤0.005) foraging and roosting sites in mature or old-growth stands within home ranges [180]. In coniferous forests of the western Cascade Range of southwestern Washington, there was significantly (P<0.02) more forest 130 years old or older in 500-acre (200 ha) areas around nest sites compared to random sites [161]. In a western Oregon study, over 90% of roosting and nesting locations were in old-growth coniferous forests, and northern spotted owls foraged in these forests significantly (P<0.05) more than would be expected based on availability within their home ranges [62]. In coniferous forests in southwestern Oregon, 83% of northern spotted owls selected old-growth forests and used an average of 1.5 times the amount of old growth than would be expected based on availability in the landscape [42]. However, there was no difference in stand age between small nest and random plots in the eastern Cascade Range of Washington [32], and only 12% of nest stands on the eastern slope of the Cascade Range in Washington and Oregon were classified as old growth [53].
The use of younger stands by northern spotted owls is mixed. For instance, in coniferous forests of the western Cascade Range in southwestern Washington, northern spotted owls selected (P<0.02) stands less than 49 years old more than expected based their availability [161]. In the coastal redwood zone, forests more than 60 years old and those less than 7 years old were used in proportion to their availability, while stands from 31 to 60 years of age were used for nesting more (P≤0.039) than expected based on availability [59]. However, in mixed-evergreen and Klamath montane forests of northwestern California, pole timber and seedling-sapling stands less than about 35 years old were not used for nesting. Roosting individuals used the pole timber stands, but less (P<0.001) than would be expected based on availability [23]. In mixed-evergreen forests of northwestern California, summer foraging and roosting sites occurred significantly (P≤0.05) less often in young- to intermediate-aged stands than expected based on availability [180].
Greater amounts of mature and old-growth forests have been associated with improved northern spotted owl persistence and reproductive output in some areas. For instance, persistence on territories in southwestern Oregon was significantly (P<0.03) higher with increased amounts of forests more than 120 years old in the 8,870-acre (3,590 ha) area around the nest site [18]. The average number of fledglings per site was significantly (P<0.05) higher in 990-acre (400 ha) areas of western Washington, western Oregon, and northwestern California with more than 60% mature (>80 years) forest than in areas with 20% or less mature forest [19].
In other locations the relationship between forest age and northern spotted owl reproductive success has been ambiguous. In forests of the northern California Coast Ranges, greater reproductive success was negatively associated with clearcuts less than 6 years old at 2 scales, positively associated with 21- to 40-year-old forest at 3 scales, and negatively associated with 61- to 80-year-old forest at all 5 scales investigated [188]. See [188] for more detail regarding forest age class selection at different scales. In the eastern Cascade Range of Washington, the amount of late-successional forest (dominated by trees with >25-inch (64 cm) diameter at breast height (DBH)) within a 500-acre (200 ha) area of the nest was negatively associated with reproductive rate. Previously occupied territories that were not used in the last 3 years of the study had significantly less forest in seedling and sapling stages and significantly more forest dominated by pole-sized trees than territories occupied during this period (P<0.05) [104].
Northern spotted owl home range size may be influenced by the amount of old-growth habitat available. In mixed-conifer and Douglas-fir forests of southwestern Oregon, home range size was inversely correlated (r=-0.83) with the percentage of old-growth habitat. The amount of old growth used ranged from 1,330 to 2,360 acres (538-955 ha), which was less variable than the 1,920- to 8,980-acre (777-3,635 ha) home range sizes [43]. A review cites a 1984 Six Rivers National Forest publication for an association between home range size and amount of old growth [82]. The northern spotted owl recovery plan also includes citations for an inverse relationship between the amount of old-growth and home range size [158].
Although much less information is available on importance of old-growth habitats to California and Mexican spotted owls, data suggest that they select old growth and/or avoid young stands. Of oak-pine and riparian forests within California spotted owl home ranges in the Sierra National Forest, 91% were old growth. In Douglas-fir, ponderosa pine, and white fir dominated forests of the Sierra Nevada, early-successional stands including clearcuts, shrublands, and plantations were avoided by foraging individuals. Only 2% of telemetry locations occurred in these cover types, while they made up 30% of the available habitat. Home ranges were also comprised of significantly (P<0.001) less of these cover types than would be expected based on availability. Increased old-growth and mature tree basal areas were also characteristic of stands occupied by California spotted owls [36]. In mixed-conifer forest of the Sierra Nevada only 13% of the vegetation within California spotted owl home ranges was classed in the greater than 21-inch DBH category. The low reproductive rate in this area suggests that it was not providing high quality habitat [213]. In northern Arizona some foraging Mexican spotted owls selected old-growth mixed-conifer and ponderosa pine forests (generally >200 years old) more than expected based on availability within the home range, while managed forests in this area were avoided [71]. According to a literature review, continuous forests used by Mexican spotted owls are typically old-growth forests [83] and territories in Arizona usually contain mature trees, as well as other features associated with mature and old-growth habitats [46].
For information on the age of trees used by spotted owls for nesting and roosting, see those sections.
Tree characteristics: Spotted owls typically select areas with large trees associated with mature and old-growth stands. In coniferous forests of Tahoe National Forest, foraging California spotted owls used large (≥21 inch (53.2 cm) DBH) tree stands significantly (P<0.005) more than expected based on availability [36]. A summary of a California spotted owl habitat study in the Tahoe National Forest reports significantly more foraging sites in stands of large (20- to 35-inch (51-89 cm) DBH) trees than expected based on availability and significantly more foraging sites in these stands than in stands of medium-sized (11- to 20-inch (28-51 cm)) trees (P<0.01). An analysis of data from the national forests of the Sierra Nevada showed that California spotted owls nested in stands of medium to large (≥24-inch (61-cm DBH)) trees more than expected based on availability [85]. Several sources note Mexican spotted owls' use of large, mature trees [46,93,205], including roosting in areas with high densities of relatively large Douglas-fir and southwestern white pine (Pinus strobiformis) in coniferous forests of New Mexico [93]. Northern spotted owl high-use sites in coniferous forests of northwestern Washington had higher densities of trees greater than 31.5 inches (80 cm) DBH and higher foliage volumes than rarely used sites (P<0.1) [156]. In low to mid-elevation conifer forests of northwestern California, nest stands had significantly more (P<0.005) conifers greater than 35 inches (90 cm) DBH than would be expected based on availability [114]. In the eastern Cascade Range of Washington, the average height of the dominant canopy in small nest plots was significantly (P≤0.02) taller than canopy height on random plots within nest stands [32]. The average size of trees greater than 39 inches (100 cm) DBH was significantly (P<0.001) larger at nest sites than random sites within home ranges in coniferous forests in Oregon and Washington. However, trees greater than 21 inches (>53.4 cm) DBH did not occur at significantly greater densities or basal areas on nest sites than on random sites. Trees from 25 to 126 feet (7.6-38.3 m) in height were significantly more dense on nesting sites than random sites (P<0.1), while density of trees in the tallest (>176 feet (53.7 m)) categories were similar on nesting and random sites [91].
Northern and California spotted owl may select habitats dominated by intermediate-sized trees in some areas. In the eastern Cascade Range of Washington, the average density of intermediate-sized (14 to 24 inches (35-60 cm) DBH) Douglas-fir trees in small northern spotted owl nest plots was significantly (P=0.03) greater than that in random plots within nest stands [32]. Basal area of trees 21 inches (53.3 cm) DBH or smaller was significantly (P<0.001) greater on small nest plots than on random sites in coniferous forests in Oregon and Washington [91]. California spotted owl home ranges contained significantly (P<0.001) more forests comprised predominantly of 11- to 21-inch (27.0-53.1 cm) DBH trees than expected based on availability in coniferous habitats within Tahoe National Forest [36]. Although the amount of variation in tree DBH was similar on northern spotted owl nest sites and random sites in coniferous forests of Oregon and Washington [91], tree size variability may be important to spotted owls in some portions of their range [55,78,114,193].
Areas with more large trees may provide higher quality habitat. In previously logged forests in the northern Coast Ranges of California, northern spotted owls with greatest reproductive success had territories with a greater density of remnant large trees than less successful individuals. These differences were significant at the 120-acre (50 ha, P=0.042) and 208-acre (114 ha, P=0.052) scales [188]. In contrast, areas with dense small trees may be associated with lower site fidelity. In the eastern Cascade Range in Washington, areas with few 5- to 7-inch (13-19 cm) DBH trees were used more often than those with more of these pole-sized trees. Abandoned territories in this study area contained significantly (P=0.049) more pole-sized stands than occupied territories [104]. In conifer forests of southern Sierra Nevada, high foliage volume above the nest was related to California spotted owl nest success [157]. In the Lassen National Forest, stands dominated by large (>24 inches (61 cm) DBH) trees were associated with greater occupancy and apparent survival, and those areas with more large remnant trees were associated with increased nest success. Nest areas dominated by small trees had lower site occupancy and reproductive output. However, pairs in stands dominated by medium-sized (12-24 inches (30-61 cm) DBH) trees had higher nest success than those in stands dominated by large trees [24].
Although a common feature of northern spotted owl nest plots, mistletoe infestation in nest plots was similar to infestation in unoccupied stands [53] and nest stands [32] in the Cascade Range. For instance, there was no significant difference (P>0.3) in mistletoe infestation rating between small northern spotted owl nest plots and nest stands, although 84% of nest plots were infested [32]. Although the mistletoe infestation rate was not investigated, medium to large (>11 inches (27.5 cm) DBH) trees at California spotted owl foraging locations had significantly (P<0.001) lower vigor than those at random locations in coniferous habitats of Tahoe National Forest [36].
For information on the characteristics of trees used by spotted owls for nesting and roosting, see those sections.
Canopy cover/Tree density: Areas used by spotted owls typically have greater than 40% canopy cover and often have more than 70% canopy cover [18,20,24,55,104,184,197,216]. Despite variation in tree height and DBH, canopy cover varied little, ranging from 88% to 95% on small nest plots in grand fir forests of eastern Washington. Areas containing 90% of telemetry locations typically had greater than 40% canopy cover in this region [55]. Review of several studies found that northern spotted owl nesting and roosting sites in the California Klamath and Coast regions had more than 80% cover [18]. Canopy cover within 82 feet (25 m) of California spotted owl nest sites in foothill riparian and oak woodlands in the southern Sierra Nevada averaged 86% [184]; and in white fir-mixed-conifer forests in northeastern California, canopy cover was typically greater than 80% in a similar-sized area around nest sites [24]. According to a review, cover of vegetation above 7 feet is typically more than 70% at California spotted owl nest sites, although canopy cover as low as 30% to 40% has been observed at higher elevations [197]. Although a review notes the occurrence of Mexican spotted owls in sparsely vegetated habitats [83], other reviews state that they typically occupy stands with more than 40% [20] or 60% canopy cover [216]. In ponderosa pine-Gambel oak forests of Arizona, canopy cover was typically greater than 40% with only 2 roost stands having canopy cover from 25% to 40%. In addition, 75% of roost stands had more than 60% canopy cover [74].
Spotted owls may select areas with high canopy cover [36,70,71,80,91,205]. In coniferous forests in Oregon and Washington, canopy cover in small plots near northern spotted owl nests was significantly (P<0.095) greater than at random plots within spotted owl home ranges. However, when data were not pooled across study areas, only nest sites in the Olympic and Cascade regions had significantly (P<0.004) greater canopy cover than random plots [91]. Six California spotted owls in coniferous habitats of the Tahoe National Forest used stands with more than 40% canopy cover significantly more, and stands with less than 40% canopy cover significantly less than expected based on availability (P<0.005). In addition, California spotted owl home ranges were comprised of significantly (P<0.02) more forest with more than 70% canopy cover than expected [36]. California spotted owls in coniferous forests in the Sierra Nevada consistently selected areas with high canopy cover [213]. According to a review, roost sites in the Sierra Nevada and in southern California had significantly higher canopy cover than random sites. However, canopy cover at nest sites was not significantly higher than random sites in all study areas [85]. In northern Arizona, Mexican spotted owl nest stands averaged 75.2% canopy cover, which was significantly (P<0.001) higher than the 53.8% average canopy cover in random stands [70]. In a ponderosa pine-Gambel oak forest in Arizona, stands with more than 60% canopy cover were used for roosting and foraging in both the breeding and nonbreeding seasons more than expected based on availability. Stands used for roosting had significantly (P≤0.03) greater canopy cover than stands that were not used for roosting [76]. Other studies that have found greater canopy cover in areas used by Mexican spotted owls than random areas are summarized in [20,78].
Spotted owls may also select habitats with high live tree basal areas. The average live tree basal area on small plots at and around northern spotted owl nests in the eastern Cascade Range of Washington was significantly (P=0.09) greater than that of random plots within nest stands [32]. In forests of the northern Coast Ranges in California, there was a significantly greater proportion of the >69 m²/ha basal area category within 17 acres (7 ha), and significantly lower proportions of the less than 23 m²/ha basal area class within 17 acres (7 ha) and 124 acres (50 ha) of owl sites compared to random sites (P<0.05) [188]. A review notes that California spotted owls in the Sierra Nevada and in southern California nested and roosted in areas with greater average conifer and total live basal areas than random locations [85]. In northern Arizona, basal area in Mexican spotted owl nest stands averaged 37.9 m²/ha, which was significantly (P<0.001) greater than the average 25.4 m²/ha in random stands [70]. According to a Mexican spotted owl habitat analysis in a review, the basal area at Mexican spotted owl nests in the Basin and Range East region was significantly (P=0.0121) greater than in random forest stands, and in the upper Gila region the basal area at nests was significantly (P<0.0001) greater than within nest and random forest stands [78]. In Douglas-fir/hardwood forests of northwestern California, the smaller northern spotted owl males foraged in stands with higher tree density than stands used by the larger females [180].
Investigations of the impact of canopy cover and basal area on spotted owl habitat have found different relationships [24,33,99,121,188,210]. In forests in the northern Coast Ranges of California, activity centers of northern spotted owl pairs in the upper 50th percentile of reproductive success had higher proportions of the 23 to 69 m²/ha basal area classes and lower proportions of the >69 m²/ha basal area category. Many of these relationships were significant (P≤0.053) at scales from 124 to 983 acres (50-398 ha) [188]. In young forests of mainly western hemlock and Sitka spruce on the Olympic peninsula, sites used multiple times had an average canopy cover of 85.6%. Although high, this was significantly (P=0.03) lower than canopy cover on unused sites or sites within northern spotted owls' home ranges that had been used once [33]. In coniferous forests in the Sierra Nevada, there were weak but significant correlations (0.29<r<0.37, P≤0.04) between canopy cover and California spotted owl reproduction. The amount of forests with 0 to 39% canopy cover was negatively correlated, and the amount of forest with more than 40% cover was positively correlated with reproductive output at the 3 scales investigated (178-1,063 acres (72 ha-430 ha)). There was typically about 10% more habitat with canopy cover >50% on sites that consistently produced young compared to unproductive sites [99]. This difference was due to increased frequency of nesting pairs with increasing canopy cover, not higher reproduction by nesting pairs [121]. Geographic variability in spotted owl prey species may partially explain these differences and is discussed further in the foraging habitat section.
For information on canopy cover and tree density at spotted owl nesting and roosting locations, see those sections.
Cover type/Dominant tree species: Although some northern and California spotted owls select and avoid certain cover types, trends related to cover type or species composition are not consistent. The average Douglas-fir basal area of small plots on and around nests was significantly (P=0.02) greater, and the average basal area of lodgepole pine (Pinus contorta), Engelmann spruce (Picea engelmannii), and western white pine (Pinus monticola) was significantly (P=0.02) less than in random plots within nest stands in the eastern Cascade Range of Washington [32]. One study found that northern spotted owls in areas of northwestern California avoided hardwood stands [23,51]. However, Folliard and others [58] found no evidence for selection of cover types in northwestern California, with northern spotted owl nests occurring in cover types ranging from purely coniferous forests to those dominated by hardwoods. Northern spotted owls in northwestern California selected forests with understories comprised of oak [114]. Community type was rarely included in models for predicting spotted owl presence in old-growth forests of Olympic National Park. This indicates that habitat structure is more influential than species composition in northern spotted owl habitat selection [140]. According to a review of California spotted owl habitat, nests of California spotted owl were observed in mixed-conifer stands more than expected based on availability in national forests of the Sierra Nevada [85]. However, in middle elevation mixed-conifer forest of the central Sierra Nevada, cover types were not significantly different between 1,129-acre (457 ha) plots on California spotted owl activity centers and random areas. The authors also summarize evidence that California spotted owls occur in a wider range of cover types than northern spotted owls [143].
Mexican spotted owls apparently respond more consistently to cover type than northern and California spotted owls. Roost sites in southern New Mexico occurred primarily in mixed-conifer forest, and owls selected mixed-conifer stands for roosting in all but 1 season/site combination [77]. Selection of mixed-conifer habitats was also observed in the Black Range and San Mateo Mountains of New Mexico [93] and in the upper Gila Mountains region of Arizona [78]. In Arizona, the majority of foraging and roosting occurred in ponderosa pine-Gambel oak forests. Foraging and roosting stands in pine-oak areas also had generally higher Gambel oak density and basal areas than unused stands [76]. Similar trends have been observed in the Upper Gila Recovery Unit [76] and in the Black Range and San Mateo Mountains [93]. Gambel oak's contribution to canopy cover, canopy layering, and prey habitat in mixed-conifer-hardwood and pine-oak communities likely influences its selection by Mexican spotted owls [76]. A review notes selection of white fir in the Basin and Range East region and several firs (Abies spp.) in Utah. There was significantly more Douglas-fir, Gambel oak, and limber pine (Pinus flexilis) on roost sites than random sites in the upper Gila Mountains region [78]. Roosting Mexican spotted owls may have selected areas with greater southwestern white pine densities in southwestern New Mexico [93]. Ponderosa pine and pinyon-juniper woodland were avoided by Mexican spotted owl in southern New Mexico [77,93]. Ponderosa pine was also more abundant in random stands than in nest stands in the Upper Gila Mountains and Basin and Range East recovery units, according to a review [78]. In southern New Mexico, Mexican spotted owls had greater survival, increased fecundity, and smaller home ranges in a mesic area dominated by mixed-conifer forest than in a xeric area dominated by ponderosa pine forests and Colorado pinyon-alligator juniper woodlands [77].
For information on tree species used for spotted owl nesting and roosting, see those sections.
Canopy layering: Northern spotted owls are strongly associated with multilayered forests. An uneven-aged, multiple-layered canopy is consistently included in descriptions of northern spotted owl habitat [42,55,62,114,134,180]. Increased canopy layering was a significant predictor of northern spotted owl presence in Olympic National Park [140]. In coniferous forests of northwestern Washington, the height class diversity, a measure of canopy layering, was significantly (P<0.1) greater in stands with more than 10% of telemetry locations compared to stands that were used less often [156]. In coniferous forests on the eastern slope of the Cascade Range in Oregon and Washington, nest sites had more sapling, pole, and large trees, while unoccupied stands within 2,953 feet (900 m) of the nest site had more medium-sized trees. This indicates a more layered canopy in occupied stands [53]. However, the average number of canopy layers in small northern spotted owl nest plots was not significantly different from that on random sites within nest stands in the eastern Cascade Range of Washington [32].
California spotted owls also used multilayered forests. Foraging sites of 6 California spotted owls had significantly (P<0.001) more vegetation layers than random locations in coniferous stands in the Tahoe National Forest [36]. In foothill riparian and oak woodlands in the southern Sierra Nevada, California spotted owl nest sites occurred in areas with multilayered canopies. The subcanopy was a dense (743 stems/acre) layer of trees less than 5 inches (13 cm) DBH that averaged 13 feet (4 m) tall. The major canopy layer was moderately dense (129 stems/acre) and was comprised of trees 5 to 30 inches (13-76 cm) DBH and 31 to 60 feet (9-18 m) tall. The upper canopy was sparse (1 stem/acre) and comprised of trees 31 to 60 inch (79-152 cm) DBH and 68 to 102 feet (21-31 m) tall [184]. Reviews also note California spotted owl's use of forests with multilayered canopies [85,197] and suggest the possible importance of vertical layers in providing perching sites, various microclimates, and higher prey species diversity [197]. See the FEIS review of tanoak (Lithocarpus densiflorus) for a concise summary of the importance of a hardwood subcanopy for northern and California spotted owls.
The Mexican spotted owl recovery plan [78,193] and a literature review of their habitat requirements [216] note the use of multilayered forests by Mexican spotted owls. In Saguaro National Park, small plots on Mexican spotted owl roost sites had significantly (P=0.049) more vegetation layers than random plots [205].
Understory: Areas occupied by spotted owls have varied understories. Tall shrub cover was significantly (P<0.098) greater on small plots around nest sites than on random sites within home ranges in coniferous forests in Oregon and Washington [91]. Anecdotal evidence suggests that shrubs in a shelterwood cut provided northern spotted owl foraging habitat [82]. However, shrub and herb cover were lower in areas frequently used for foraging and roosting than on infrequently used sites in mixed-evergreen forests of northwestern California [180]. In forests in the western Cascade Range, stands frequently used by northern spotted owls for foraging had lower herb and shrub cover than random stands in the same age class [105]. In foothill riparian and oak woodlands in the southern Sierra Nevada, California spotted owl nest sites had little vegetative ground cover, with an average of 20% small shrub cover and 21% grass cover. An average of 60% of the forest floor was covered by small litter [184]. A review of Mexican spotted owl habitat data notes that Rinkevich and Gutierrez (1996) found higher litter cover on roost sites than on random sites in southern Utah [78]. A review of the impacts of ungulate grazing on native species of the southwest suggests that a variable shrub layer provides both habitat for prey species and foraging habitat for Mexican spotted owls [216].
Snags: In some areas, northern spotted owls select habitats with more snags. For instance, in coniferous forests Oregon and Washington, basal area of relatively sturdy snags was significantly (P<0.001) greater in small plots near nests than in random sites within home ranges [91]. In young coniferous forests on the Olympic Peninsula in Washington, sites used repeatedly by northern spotted owls had significantly (P=0.0007) more snags greater than 20 inches (51 cm) DBH than random or single-use sites [33]. Similarly, in coniferous forest of Oregon, snags larger than 16 inches (40 cm) DBH were more abundant in stands used frequently for foraging than in random sites in forests of the same age class [105]. Snag volume was also greater (P<0.1) on northern spotted owl high use sites than on less often used sites in coniferous forests of northwestern Washington. Stands with snag basal areas less than 142.1 m³/ha were typically used less frequently for foraging [156]. Basal area of highly decayed snags (P<0.001) and density of small snags (P=0.08) were greater on small nest plots in the eastern Cascade Range of Washington than on random plots within nest stands. However, basal area of hard snags was significantly lower (P<0.01) on nest plots. Densities of larger snag size classes, and basal areas of snags in other decay classes were similar on nest sites and random sites within nest stands [32]. In coniferous forests of Washington and Oregon, the abundance of snags was similar on nest sites and unoccupied stands within a 0.6-mile (0.9 km) radius [53]. In old-growth coniferous forests of Olympic National Park, snag diameter was a significant predictor of owl presence. However, the relationship changed with study area, with larger snags associated with owl presence on the eastern side of the Bailey Range, and smaller snags associated with owl presence on the western side [140].
California and Mexican spotted owls may select habitats with abundant snags. According to a review, snag basal area at California spotted owl nest and roost sites was significantly (P<0.05) greater than at random sites [85]. Foraging locations in coniferous forests of the north-central Sierra Nevada also had significantly (P<0.001) greater snag basal area than random locations [36]. In foothill riparian and oak woodlands in the southern Sierra Nevada, California spotted owl nest sites averaged 14 snags/acre [184]. A review and analysis of habitat data tentatively recommends snag basal areas in Sierra mixed-conifer forests of 15 to 30 ft²/acre in foraging habitat and 30 to 55 ft²/acre in roosting and nesting habitat. It also suggests that snags greater than 15 inches (38 cm) DBH comprise 7 to 17 ft²/acre of foraging stands and 20 to 30 ft²/acre of nesting and roosting stands [85]. Mexican spotted owl roosting and foraging sites in coniferous forests of Arizona had significantly (P<0.001) greater snag densities and basal areas than random sites. See the table below for details [71]. A review reports significantly (P<0.0001) greater average snag densities in areas close to Mexican spotted owl nest sites in the Upper Gila Mountains Recovery Unit in Arizona and New Mexico. In small nest plots, snag density averaged 63.9/ha. Nest stands averaged 44.0 snags/ha and random stands averaged 17.6 snags/ha. Snag basal area was also significantly (P=0.0003) higher in nest stands than in random stands. Occupied canyons in Zion National Park also had higher snag basal areas than canyons where Mexican spotted owls were not detected [78].
Snag densities and basal areas in Mexican spotted owl roosting and foraging areas and at random areas in coniferous forests of Arizona [71] Density (snags/ha) Basal Area (m²/ha) Roosting 97.3 8.9 Foraging 55.1 6.4 Random 22.5 2.4
Coarse woody debris: Northern spotted owl may select habitats with more coarse woody debris in some areas, possibly due to the apparent dependence of northern flying squirrels (Glaucomys sabrinus) and other spotted owl prey species on coarse woody debris for cover and truffle production [197]. In coniferous forests in Oregon and Washington, the volume of highly decayed logs was significantly (P≤0.025) greater in small plots near northern spotted owl nests than in random sites within home ranges [91]. The volume of downed logs with diameters greater than 20 inches (50 cm) was significantly (P=0.0002) greater on nesting and frequently used foraging sites than on random sites in coniferous forest in the western Cascade Range in Oregon [105]. However, in the eastern Cascade Range of Washington, the average volume of coarse woody debris in small nest plots was similar to that in random plots within nest stands [32]. Log abundance was also similar in nest stands and unoccupied stands within a 0.6-mile (0.9 km) radius of nest sites in coniferous forests on the eastern slope of the Cascade Range in Washington and Oregon [53].
Coarse woody debris is likely an important feature of Mexican and California spotted owl habitats. In coniferous forests of Arizona there were significantly (P<0.001) more logs with greater than 12-inch (30.5 cm) diameters and lengths of 10 or more feet (3 m) on sites used by 8 Mexican spotted owls for roosting and foraging than on random sites [71]. According to a review, the average amount of coarse woody debris was 97.8 m³/ha at nest sites and 94.7 m³/ha in nest stands in the Upper Gila Mountains Recovery Unit of Arizona and New Mexico. This was significantly greater (P=0.0006) than the average of 54.6 m³/ha in random stands. Nest sites in the Basin and Range East Recovery Unit in New Mexico also had higher (P=0.0061) log volumes than random stands [78]. Six California spotted owls selected 0.1-acre (0.04 ha) foraging plots with significantly (P<0.001) more cover of coarse woody debris than on random sites in the Tahoe National Forest [36]. Winter roost sites in the Eldorado National Forest also had significantly (P<0.05) more coarse woody debris than random sites, according to a literature review and analysis [85]. Another review states that coarse woody debris is more influential in coniferous forests than in riparian/hardwood habitats, due to the role of hypogeous fungi in supporting spotted owl prey in coniferous forests [197]. In foothill riparian and oak woodlands in the southern Sierra Nevada, the amount of coarse woody debris greater than 10 inches (25 cm) in diameter ranged from 156 to 331 ft³/acre [184]. A review and analysis of habitat data gives tentative recommendations for coarse woody debris density in Sierran mixed-conifer forests of 10 to 15 tons/acre in nesting, roosting, and foraging stands. It also recommends that coarse woody debris be >11 inches in diameter. The lower end of observed coarse woody debris densities and larger pieces of coarse woody debris were recommended to reduce the risk of fire in spotted owl habitats [85]. See the HABITAT RELATED FIRE EFFECTS and Fire Ecology sections for more detailed information on the risk of fire in spotted owl habitat.
Foraging Habitat: Although similar to nesting and roosting habitat requirements, foraging habitat requirements are likely less strict. Like nesting and roosting sites, foraging areas are generally older [43,76,180], have higher canopy covers [36,85], greater tree densities [36,78,85], and more snags and coarse woody debris [78,85,105] than random sites. In the Coast Ranges of southern Oregon, northern spotted owls foraged in forests older than expected [43]. In northwestern California, foraging sites were older than sites that were infrequently used [180]. However, California [85] and northern [105] spotted owls may use younger forests for foraging than for nesting or roosting. For instance, stands as young as 27 years old were used by foraging northern spotted owls in coniferous forests of the western Cascade Range in Oregon [105]. Although Mexican spotted owls foraged in old stands and stands with more than 60% canopy cover in ponderosa pine-Gambel oak forests in Arizona, canopy cover on foraging sites was lower than on roosting sites [76]. California spotted owls did not select areas with greater canopy cover at the patch scale (≥5 acres, 2 ha) as consistently for foraging as they did for roosting in the Sierra Nevada [213]. Log volume and snag basal area and density were similar on foraging and roosting sites on sites in Arizona and New Mexico, but stands with less than 60 ft²/acre basal area, less than 25% canopy cover, and very high Gambel oak densities were used for foraging and not for breeding-season roosts [74]. Mexican spotted owl foraging sites also differed from roosting sites in study areas near Flagstaff and Alpine, Arizona, with large coarse woody debris, density of snags, tree density, and canopy cover significantly (P<0.001) higher on roosting sites [71]. Some sources suggest that foraging habitat has more open area under the canopy to allow for spotted owl flight [81,92,197]. Increased habitat heterogeneity (See Landscape/Scale Effects) may also be important to foraging spotted owls by providing more varied prey [59,65,200].
Predominant prey (See Food Habits) is likely to contribute to differences in spotted owl habitat selection across their range, since spotted owl prey species occur in a variety of habitats. For instance, in coniferous forests of southwestern Oregon and the Olympic Peninsula in Washington, flying squirrel densities in old-growth forests were typically about twice those in younger stands [42], while dusky-footed woodrats (Neotoma fuscipes) in northwestern California tend to occur in early-successional shrublands [171]. Young forest stands did not positively affect California spotted owl reproduction or site occupancy in a white fir-mixed-conifer study area on the Lassen National Forest [24]. According to a personal communication cited in a review, northern flying squirrel was the major prey item in this area [197]. In contrast, in areas where dusky-footed woodrats are a substantial component of the diet, such as in some areas of northwestern California [17,59], some interspersion of younger or more open stands with mature and old-growth forests is likely beneficial [59,64,188]. Northern spotted owl home range size in Douglas-fir and mixed-conifer study areas in southwestern Oregon and northwestern California was inversely correlated with proportion of woodrats (Neotoma spp.) in the diet (r=-0.8, P<0.005). It is suggested that the larger mass and tendency of woodrats to occur at higher densities than flying squirrels makes for more efficient foraging and allows for smaller home ranges [211]. The 1,700-ha area of old growth used by northern spotted owls in western hemlock forests of Washington was much greater than the 500-ha area of old growth used in mixed-conifer and Douglas-fir forests of Oregon, possibly due to the greater prey biomass density in the Oregon study area (388 g/ha) than the Washington study area (61 g/ha) [42]. Reviews (e.g. [14,207]) and more detailed information (e.g., [25,42,124,171,202]) on spotted owl prey species' availability and the habitats they select are available.
Nonbreeding habitat: Although habitat of northern and California spotted owls during the nonbreeding season is generally similar to breeding-season habitat, some differences have been reported. A review reports that northern spotted owl displays less consistent habitat selection in the nonbreeding than in the breeding season [18]. According to another review, wintering habitats of northern spotted owls in the Klamath physiographic region tend to have more and larger hardwoods, such as tanoak. California spotted owls use habitats with similar canopy closure in the breeding and nonbreeding seasons, but nonbreeding habitat is generally lower in elevation and less structurally complex [83]. Preliminary data cited in a review suggest wintering habitat occurs at lower elevations than breeding habitat in the Sierra Nevada [120].
Mexican spotted owls winter in lower elevation habitats that are more open and shrubby than breeding season habitats, according to reviews [78,83]. Based on a review of radio-telemetry data from studies in Arizona, New Mexico, Colorado, and Utah, roost stands used by Mexican spotted owls in the nonbreeding season had less basal area of live trees and hardwoods and lower canopy cover than breeding season roost stands. In addition, there were more forests with more than 60% canopy cover in breeding season ranges than in winter ranges [73]. In ponderosa pine-Gambel oak forests of Arizona, roosting sites occurred in more variable habitats in the nonbreeding season than in the breeding season. Nonbreeding season roost sites occurred in areas with basal areas less than 60 ft²/acres, less than 25% canopy cover, and very high oak densities, while these habitats were not used for roosting during the breeding season [74].
Dispersal habitat: Spotted owls may disperse through vegetation that is more open than typical habitat [60,78,83,93,129,206]. Forsman and others [60] and reviews [47,83] note northern spotted owl dispersal through cover types generally considered unsuitable habitat, including relatively open areas. For instance, dispersal of northern spotted owls through fragmented landscapes of Washington and Oregon was noted in a review [129]. Another review on the status and trends of northern spotted owls and their habitat describes northern spotted owl dispersal habitat as forests with average tree diameters greater than 11 inches (28 cm), greater than 40% canopy cover, and open areas under the canopy for flight [47]. Mexican spotted owls dispersed through pinyon-juniper woodlands, mountain shrubland, desert scrub, and subalpine and mixed-conifer forests in southern Utah [206]. In southwestern New Mexico, 2 Mexican spotted owls used juniper savanna and grasslands during dispersal [93].
Landscape/Scale effects: Northern spotted owls apparently select large habitat patches. On Washington's Olympic Peninsula, the average area of habitat patches in 8,038-acre (3,253-ha ) circles centered on pair locations was significantly (P<0.01) larger than in circles centered on random sites [126]. In Douglas-fir forests of northern California, frequency of occurrence increased as stand size increased (P<0.1) from <25 acres (10 ha) to 52 to 247 acres (21-100 ha). Frequency of occurrence was also positively associated with continuous amount of hardwoods adjacent to the nest stand (P<0.05) [169]. In a model based on data from Douglas-fir/hardwood and mixed-conifer forests of California, survival was positively associated with large core habitat patches in an area of approximately 400 acres (160 ha) around territory centers. However, reproduction was negatively associated with the area of core habitat patches [64].
Less is known about California and Mexican spotted owls' relationships to patch size. In mixed-conifer forests of the central Sierra Nevada, 1,129-acre (457 ha) areas used by California spotted owls were not comprised of significantly (P=0.547) larger habitat patches than random areas [143]. However, in a model of California spotted owl dynamics and reserve design, persistence was greater when reserved areas were arranged in fewer larger patches than when they occurred in more smaller patches [9]. Nest stands in study areas throughout California averaged 100 acres (40.5 ha) [85]. A review suggests that during winter, smaller forest patches may be used than in the breeding season [83]. According to a literature review and panel of experts convened to investigate effects of grazing on native species of the southwest, Mexican spotted owls can use "small patches of dense nesting habitat" [216].
Although the results are mixed, spotted owls may be negatively impacted by fragmentation and isolation. This has most often been observed at scales from 500 (200 ha) to several thousand acres. In Douglas-fir/hardwood and mixed-conifer forests of northwestern California, northern spotted owl nest sites were significantly (P<0.01) less fragmented than random plots at approximately 500-acre (200 ha) and 1,100-acre (450 ha) scales. At approximately 2,000- to 10,100-acre (800-4,100 ha) scales, fragmentation on northern spotted owl nest sites and random sites were not significantly (P>0.05) different [100]. A similar pattern was observed in Douglas-fir forests of northern California, with no significant associations between northern spotted owl frequency and percent clearcut or total edge at large (2,500 acre (1,000 ha)) scales. Northern spotted owl frequency also had no significant association with distance to a clearcut or total length of edge within the plot at small scales (25 acre (10 ha)). However, at the stand scale, northern spotted owl frequency was negatively associated with a measure of the percentage of the stand's perimeter that bordered clearcuts (P<0.1) [169]. Although several measures of fragmentation did not differ (P≥0.14) between northern spotted owl and random areas at a large scale in the Olympic Peninsula in Washington, a measure of northern spotted owl habitat isolation was significantly (P<0.01) lower in 8,038-acre (3,253-ha) circles centered on pair locations than in random circles [126]. Effects at large scales were observed in mixed-conifer and Douglas-fir landscapes in southwestern Oregon. Northern spotted owls in fragmented areas had lower densities, less home range overlap between members of the same pair, higher home range overlap between members of neighboring pairs, and greater incidence of mate changes than owls in study areas with clumped habitat [42]. Results from a model of California spotted owl dynamics and reserve design suggest that a lightly harvested matrix between areas reserved for spotted owls would result in longer persistence times than a heavily harvested matrix, due to quicker colonization of unoccupied habitat [9]. A review [193] notes that persistence of Mexican spotted owl populations may also be beneficially impacted by higher connectivity between patches, although another review [216] concludes habitat connectivity is not crucial for Mexican spotted owls based on dispersal of more than 19 to 25 miles (30-40 km) through unsuitable habitat.
In some areas, northern and Mexican spotted owls may benefit from habitat heterogeneity, while available information suggests California spotted owls in the Sierra Nevada are negatively associated with habitat heterogeneity. In the coastal redwood zone, approximately 500-acre (200 ha) plots centered on northern spotted owl nesting sites had more (P=0.003) total edge than random plots. The positive association appeared related to the length of boundaries between different cover types or between age classes of the same cover type [59]. Modeling the role of habitat in survival and reproduction of northern spotted owls in Douglas-fir/hardwoods and mixed-conifer forests of northwestern California suggests that owls in approximately 400-acre (160 ha) areas with more edges between spotted owl habitat and other cover types have higher reproduction and, to an extent, higher survival. Overall, high quality spotted owl habitat has greater apparent heterogeneity than low-quality habitat [64]. However, northern spotted owl in the same study area had marginally (P=0.058) less seral stage heterogeneity than random plots at a scale of approximately 500 acres (200 ha) [100]. A review states that Mexican spotted owls require heterogeneous habitat with edge and notes the use of mixed-conifer forest edges and montane meadows for foraging [216]. In contrast, areas used by California spotted owl had significantly (P=0.0002) lower habitat heterogeneity than random locations at the 1,129-acre (457 ha) scale in middle elevation mixed-conifer forest of the central Sierra Nevada [143].
Differences in the abundance and diversity of prey species across the spotted owl's range are often cited to account for the variation in area, edge, and heterogeneity of the habitat used [42,59,64,211]. For more information on the role of prey in habitat requirements see the Foraging habitat section. Differences in landscape character across their range may also explain some of the variation in spotted owl habitat use [18].
Spotted owls may be more selective of habitat in areas closer to their nest sites. In mixed-conifer and ponderosa pine-Gambel oak forests of north-central Arizona, Mexican spotted owls were most selective within about 500 acres (200 ha) surrounding the nest site [80]. In California's Klamath physiographic region, differences in habitat selected by northern spotted owls and those available at random were greatest at the 200-ha scale compared to the 500- and 900-ha scales. However, sample sizes were larger at the 200-ha scale [210].
Landscape-level requirements have led to recommendations to reserve closely spaced clusters of spotted owl habitat [116,129]. Modeling the response of northern spotted owls to reserves of varying size and spacing led to the conclusion that patches should be large enough to support 20 to 25 territories. Above this size, connectivity and extent of the reserve are just as important as patch size to the probability of population persistence [116]. The importance of connectivity for dispersal and mate-finding was noted by [9,154]. Arrangement of spotted owl habitat using an optimization technique was investigated by [94]. Several studies suggest that reserves be placed in a pattern that minimizes the risk of fire [15,38] or allows for mitigation of fuel reduction effects [167]. Minimizing the risk of fire in northern spotted owl habitat, including landscape-scale recommendations, is discussed in detail in the HABITAT RELATED FIRE EFFECTS and Fire Ecology sections.
Home range/Density: The following home range sizes are calculated using a variety of field and statistical methods.
Spotted owl home ranges are generally large, but sizes are variable. The average home range size of northern spotted owl pairs varies from 1,030 acres (417 ha) in coniferous forests of Oregon [105] to 14,169 acres (5,734 ha) on Washington's Olympic Peninsula [126]. In riparian hardwood forests of the Sierra National Forest, California spotted owl had comparatively small home ranges, varying from 661 to 985 acres (267-399 ha), while those in mixed pine, white fir, and California red fir forests of the Lassen National Forest had home ranges varying from 7,061 to 12,473 acres (2,857-5,048 ha) [213]. A review notes median California spotted owl pair home range sizes of up to 18,706 acres (7,570 ha) [83]. A Mexican spotted owl review includes individual home range estimates from 645 acres (261 ha) in the upper Gila Mountains to 3,672 acres (1,487 ha) on the Colorado Plateau [78]. Pair home range estimates ranged from 2,548 acres (1,031 ha) in Arizona to 2,780 (1,125 ha) in New Mexico [83]. In some cases, Mexican spotted owls can spend a substantial portion of their time in a small portion of their home range [76,206]. For example, in riparian areas, pinyon-juniper, and mixed-conifer woodlands of southern Utah, 70% of radio locations occurred within an area averaging 689 acres (279 ha), which is less than one-third of the 2,179-acre (882 ha) area that was occupied by 95% of radio locations [206].
Spotted owls typically have smaller home ranges in the breeding season than in the nonbreeding season [43,62,73,76,206]. For example, in the western hemlock zone of the southern Oregon Coast Ranges, average northern spotted owl breeding season home range was 1,497 acres (606 ha), compared to an average nonbreeding season home range size of 3,509 acres (1,420 ha) [43]. Radio-telemetry data on Mexican spotted owls from studies in Arizona, New Mexico, Colorado, and Utah resulted in estimates of breeding season home ranges that varied from 563 to 1,250 acres (228-506 ha), while nonbreeding season home range estimates varied from 902 to 2,540 acres (365-1,028 ha) [73]. In a ponderosa pine-Gambel oak forest of Arizona, nonbreeding home range size were significantly (P=0.008) larger than breeding home ranges [76]. However, in the Sierra National Forest there was an interaction between cover type and season on California spotted owl territory size, such that nonbreeding territories were larger than breeding territories in mixed-conifer forest but smaller in oak/pine woodland [213].
Other sources of variability in home range size include habitat requirements and prey availability. Spotted owls likely require a certain amount of old-growth forests (see Succession). A literature review states that a pair of northern spotted owls uses an average of 2,000 to 2,500 acres of old-growth forests [41]. In areas where this resource is clumped and abundant, spotted owl home ranges are generally smaller. For example, in coniferous forests in southwestern Oregon, northern spotted owl home range size varied from 1,320 acres (533 ha) in clumped mixed-conifer forest to 7,190 acres (2,908 ha) in fragmented Douglas-fir forest [42]. Carey and others [43] as well as reviews [82,158] note the inverse relationship between home range sizes and the amount of old growth within them. Spotted owls also require less area in certain cover types [77,211,213]. For instance, Mexican spotted owl home ranges in the Sacramento Mountains of New Mexico dominated by mixed-conifer forests were significantly (P<0.04) smaller than those dominated by ponderosa pine forests and pinyon-juniper woodland. In addition, home range size was inversely related (P≤0.003) to the amount of mixed-conifer forest in the home range [77]. The variation between different habitat types may be related to the availability of spotted owl prey [42,158]. For instance, the low numbers and inconsistent availability of prey on Washington's Olympic peninsula was suggested as a cause for the large northern spotted owl home ranges in this region [158]. See the Foraging habitat section for more information on the role of spotted owl prey species in habitat use.
Female spotted owls may have larger home ranges than males. In mixed-evergreen forests of northwestern California, female northern spotted owl summer home ranges averaged 1,329 acres (538 ha), which was larger than the average male home range of 835 acres (338 ha) [180]. In pine-oak forest in Arizona, female Mexican spotted owls had considerably larger (P=0.073) home ranges than males [76]. However, there was no difference in the size of male and female California spotted owl home ranges in mixed-conifer and oak-pine woodlands of the Sierra National Forest and mixed pine, red and white fir in the northern Sierra Nevada [213].
The amount of overlap between members of pairs and between adjacent owls varies with season. In the western hemlock zone of the southern Oregon Coast Ranges, home range overlap of northern spotted owl pairs was 74% to 97% during the breeding season and 64% to 91% during the nonbreeding season [43]. In the Coast and Cascade ranges in western Oregon home range overlap averaged 68% between members of a pair and 12% for individuals in neighboring territories [62]. Home range overlap between members of California spotted owl pairs in coniferous forests of the Tahoe National Forest was 47% to 63% [36].
Spotted owl densities also vary with habitat and location. In mixed-conifer and Douglas-fir forests in southwestern Oregon, density of resident northern spotted owl pairs ranged from 0.046 pairs/km² in fragmented Douglas-fir forest to 0.190 pairs/km² in clumped mixed-conifer habitat [42]. In coastal redwood forests, Douglas-fir forests, and oak woodlands of northwestern California, average density of northern spotted owls was 0.209 owls/km², but varied from 0.092 to 0.351 owls/km² across subregions. Ecological densities, defined as owls per area of suitable habitat, differed significantly (P<0.001) across the subregions and varied from 0.373 to 1.049 owls/km² [51]. According to a review of distribution and abundance information in the Mexican spotted owl recovery plan, Mexican spotted owl densities averaged 0.275 owls/km² in mixed-conifer forests, 0.08 owls/km² in pine forests, and 0.022 owls/km² in pinyon-juniper woodlands of the Sacramento Mountains New Mexico [201].