Review of the Science Directly Related to the Effects of Barred Owls on Spotted Owls
This blog post was based on a report prepared for the Western Klamath Restoration Partnership (WKRP) and can be downloaded here. Addiitonally, this blog post served as the basis for a paper we published in the Journal of Wildlife Management.
Barred Owls (Strix varia) and Spotted Owls (Strix occidentalis) are both large, forest-dwelling predators. Although both Barred and Spotted Owls rely on forested landscapes, the Barred Owl exhibits a more diverse diet – consuming mammals, birds, fish, amphibians, and invertebrates (Mazur and James 2000) – compared to Spotted Owls that specialize on mammalian prey (Gutiérrez et al. 1995, Hamer et al. 2001, Wiens et al. 2014). Unlike the Barred Owl, the Spotted Owl is emblematic of controversy surrounding their dwindling numbers, protected status, and reliance on economically-valuable timberlands (Keane 2017). Because Barred Owls exert a disproportionately negative influence on Spotted Owl fitness where the two species co-occur, the continuing expansion of Barred Owls into the range of the Northern and California Spotted Owl has presented emergent and difficult challenges for wildlife managers.
In this blog, we aimed to summarize the existing published research on Barred Owls within the range of the Spotted Owl as well as identify future research priorities. Broadly, we provide synthesized information for researchers and managers to help develop strategies focused on mitigating the deleterious effects of Barred Owl expansion in western North America.
Westward Expansion of the Barred Owl
Before the second half of the last century, the ranges of the Spotted Owl and Barred Owl did not overlap. The Spotted Owl occupied the western edge of North America (Gutiérrez et al. 1995) and the Barred Owl occupied the eastern edge (Mazur and James 2000), where the Great Plains divided the two species. Barred Owls began to expand their range across the midwestern United States and central Canada in the early part of the last century (Livezey 2009a). There are several non-mutually exclusive hypotheses regarding what led to this westerly expansion, including: climate change resulting in warmer weather (Johnson 1994, Monahan and Hijmans 2007), changes in the Great Plains such as increased woody development in the form of tree plantings for shelterbelts and fire control (Knopf 1994, Livezey 2009a, 2009b), and other habitat manipulations (Root and Weckstein 1994).
After crossing central North America, the Barred Owl expanded its range into the Pacific Northwest where the first individual was observed in Alberta in 1912 (Boxall and Stepney 1982), British Columbia in 1943 (Grant 1966), eastern Washington 1966 (Rogers 1966), western Washington 1986 (Sharp 1989), Oregon 1974 (Taylor and Forsman 1976), northwestern California 1982 (Evens and LeValley 1982), and as far south as Marin County in California by 2002 (Jennings et al. 2011) and the Sierra Nevada Mountains in eastern California in 1991 (Dark et al. 1998). The western Barred Owl range now overlaps the range of the Northern Spotted Owl (S. o. caurina) (Livezey 2009a) and the California Spotted Owl (S. o. occidentalis).
The Barred Owl has yet to expand into the range of the California Spotted Owl in the densities seen in the Northern Spotted Owl range (Keane 2017). As of 2013, 51 Barred Owls and 27 hybrids had been detected in the Sierra Nevada range, though none in the coastal or southern California parts of their range. Barred Owl numbers remain relatively low in the California Spotted Owl range when compared to the Northern Spotted Owl range, though contemporary survey efforts for Barred Owls remain incidental (Keane 2017).
Barred Owl influence on Spotted Owl habitat use and selection
As Barred Owls expanded their range across western North America, their newly established populations in British Columbia, Washington, Oregon, and northern California continued to grow (Dark et al. 1998, Davis et al. 2016, Jones and Kroll 2016, Kelly et al. 2003, Pearson and Livezey 2003). Pervasive expansion and population growth continued until only a few isolated areas in the range of the Northern Spotted Owl remained without Barred Owls (Kroll et al. 2016). Barred Owl expansion eventually reached a critical threshold in many parts of the northern Spotted Owl range: the number of territorial Barred Owls surpassed Spotted Owl territories (Fig. 1, Lesmeister et al. 2016).
Figure 1. Proportion of spotted owl sites in which barred owls and spotted owls were detected on the Oregon Coast , 1990–2015. From Lesmeister et al. (2016).
To date, few studies have tried to estimate Barred Owl population size. In one such case, Kelly (2001) estimated there were a total of 706 Barred Owl territories in Oregon by 1998, slightly more than 20 years after the first Barred Owl was recorded in Oregon in 1974. However, Gutiérrez et al. (2004) stated that detection methods which reported cumulative detections may lead to over estimates of Barred Owl populations. On the other hand, Gutiérrez et al. (2004) also stated that most Barred Owl sightings are reported incidentally during Spotted Owl surveys, suggesting that they may be more abundant than estimated. A new modelling framework developed by Zipkin et al. (2017) has leveraged these detection-nondetection data along with count data to estimate population dynamics, abundance, and individual detection probabilities from sampling Barred Owls in the Oregon Coast Ranges. They estimated that the mean site-specific number of 0.13 territorial Barred Owls in 1995 increased to 7.5 owls in 2016, with survival probabilities of 0.86-0.93 and an increased colonization rate of 0.14 in 1996 to 0.90 in 2016. Developing and implementing Barred Owl surveys represents an important step towards fully understanding how environmental variation affects Barred Owl population growth.
[endif]--Both Barred and Spotted owls utilize old-growth forests. While four researchers found no differences between species’ use of different age classes and composition of forests (Buchanan et al. 2004, Pearson and Livezey 2007, Singleton 2015, Singleton et al. 2010), others found Spotted Owls use older forests at greater frequencies than Barred Owls. For example, Hamer et al. (2007) reported that the home ranges of Spotted Owls in Washington were negatively influenced by the lack of old forest, that is, home ranges with less old forest were larger, indicating Spotted Owls may increase their range to increase amount of old forest in their territory. This was only slightly true for Barred Owls. Also in Washington, Herter and Hicks (2000) found that Spotted Owl territories contained more mature coniferous forests than Barred Owl territories and Pearson and Livezey (2003) found the mean age of forest stand at a site-center was significantly greater for Spotted Owls (254.7 ± 76.5 yr) than for Barred Owls (228.3 ± 101.5 yr). In Oregon, Wiens et al. (2014) found that both species used patches of old (>120 yr) coniferous forest in proportions two to three times greater than available in the study area (Table 1). In northern California, Weisel (2015) researched the owls’ use of the understory components of the coastal redwood ecosystem and found that both species selected habitats with understory vegetation, hardwood trees, and close to their nest sites; however, Barred Owls were more likely to select foraging habitat that contained a greater percentage of hardwoods than Spotted Owls only up to a certain volume, when the stand becomes too dense to support successful foraging, suggesting that Spotted Owls are better able to utilize dense understory for foraging. Similarly, Irwin et al. (2017) found that Barred Owls selectively hunted for prey near streams at low elevations, often within hardwood-dominated dominated stands, but use decreased with increasing densities of small-diameter trees.
Table 1. Mean values of environmental conditions measured at foraging and roosting locations used by individual Northern Spotted Owls or Barred Owls as compared to a set of random locations plotted in the western Oregon study area, USA, 2007–2009. Forest types are expressed as the mean percentage of total foraging, roosting, or random locations. We show sample sizes (number of individual owls or random points) in parentheses. From Wiens et al. (2014).
Habitat around nest sites in the Cascades were found to be similar between the species (Buchanan et al. 2004, Pearson and Livezey 2003). However, Buchanan et al. (2004) found differences in nest trees and nest configurations selected by each species. In Washington, Spotted Owls almost always placed nests in Douglas-firs (Psuedotsuga menziesii) (9 of 10 nests), while Barred Owls used five different tree species, including three black cottonwoods (Populus trichocarpa), three Douglas-firs, and two grand firs (Abies grandis) (Buchanan et al. 2004). Most of the trees used by Spotted Owls were alive and intact (7 of 10) while eight trees used by Barred Owls had broken boles (six alive, two dead). Also, Spotted Owls primarily used platform nests consisting of clumps of branches or goshawk nests (8 of 10), while Barred Owls used cavities or sites with chimney-like structures. Allen (1987) noted that Barred Owls in the east required large, decadent trees with cavities for nesting, though Livezey (2007) found 25% used other locations such as hawk nests, tops of hollow trees, and squirrel nests throughout their range.
Barred Owls appear to have a negative effect on the Spotted Owl’s ability to utilize preferred habitat. For example, Davis et al. (2016) modeled habitat suitability in the Tyee density study area in Oregon and found a strong negative correlation (r = −0.894) between increasing trend of Spotted Owl territories with Barred Owls and the average Habitat Suitability Index at annual Spotted Owl locations. The 2013 index was significantly lower than it was in 1990 when the study began with Barred Owls occurring in low numbers. Pearson and Livezey (2003) found that Late Successional Reserves (LSRs) in Washington were getting more use by Barred Owls than Spotted Owls for whom they were set aside. They found 34% more Barred Owl sites than Spotted Owl sites in LSRs, while there were 33% more Spotted Owl than Barred Owl sites in non-reserve lands.
The greatest difference in habitat use between Spotted and Barred Owls is an apparent dissimilar use of topographies by the two species: Spotted Owls used steeper, higher-elevation sites while Barred Owls used flatter, low-elevation sites – sometimes along streams (Hamer et al. 2007, Herter and Hicks 2000, Pearson and Livezey 2003, 2007, Wiens et al. 2014) (Table 1). Pearson and Livezey (2007) found that elevation and slope were important factors in explaining densities of Spotted Owls in Washington, when combined with measures of forest quality, forest age, distance to water, and abundance and availability of prey. They posited that the persistence and higher numbers of Spotted Owls in one LSR, in spite of the invasion of Barred Owls in other LSRs, may indicate that there are local environmental factors such as elevation and slope that favor Spotted Owls over Barred Owls, and that a natural balance had been achieved in their study area which allowed the coexistence of these two species.
Influence of Barred Owls on Spotted Owl Occupancy, Survival, and Fecundity
As Barred Owl populations increase, they can either form new territories in Spotted Owl territories without displacing Spotted Owls (Wiens et al. 2014), actively displace resident Spotted Owls from their territories (Dugger et al. 2011, Kelly et al. 2003, Pearson and Livezey 2003, Sharp 1989), or reduce Spotted Owl occupancy rates (Kelly et al. 2003, Kroll et al. 2010, Pearson and Livezey 2003). When Barred Owls establish territories inside Spotted Owl territories, some Barred Owl territories may be contained entirely within Spotted Owl territories , since the estimated size of Spotted Owl territories can be more than six times the size of Barred Owl territories (Table 2) (Hamer et al. 2007, Wiens 2012, Wiens et al. 2014). Wiens et al. (2014) found that each individual Spotted Owl shared a portion of its home range, usually foraging areas, with 0-8 Barred Owls in adjacent territories (average = 2.4 Barred Owls per Spotted Owl). Spotted Owls were less likely to use an area if it was within or near a core-use area of a Barred Owl (Wiens et al. 2014). By contrast, in coastal northern California, Weisel (2015) found no significant difference in average home range size between the two species in either the breeding or non-breeding seasons, however Spotted Owl territories tended to be slightly larger (Table 2).
Table 2. Home range sizes (mean ha ± SE) for Northern Spotted Owls and Barred Owls in Washington, Oregon, and northern California. Hamer et al. (2007) estimates were calculated using 95% Adaptive Kernel Method, all others used 95% Fixed Kernel Method.
One of the earliest incidents of “displacement” of a Spotted Owl by a Barred Owl was recorded in Washington state by Sharp (1989). In 1985, Sharp recorded the first Barred Owls on the Olympic Peninsula. The following year, two Barred Owl pairs had apparently displaced two Spotted Owl pairs from their territories.
The presence of Barred Owls yields a significant effect on Spotted Owl colonization and extinction rates. Five of the six studies that estimated colonization probabilities found a negative effect of Barred Owls on colonization, though it was sometimes weak (Dugger et al. 2011, 2016; Kroll et al. 2010; Olson et al. 2005; Yackulic et al. 2014). Sovern et al. (2014) found no effects of Barred Owls on colonization rates. Dugger et al. (2016) found the presence of Barred Owls resulted in a lower local colonization rate in five of eleven study areas in the Northern Spotted Owl range. Also, Yackulic et al. (2014) found that two of their top four models for the Tyee study area suggested that Spotted Owls were less likely to colonize an area that was already occupied by Barred Owls, while, conversely, Barred Owls were more likely to colonize areas already occupied by Spotted Owls. Dugger et al. (2011) found that Barred Owls displaced Spotted Owls from historical breeding territories in southern Oregon, as Spotted Owl site occupancy was lower where Barred Owls were detected as compared to sites where Barred Owls were not detected (Fig. 2). Similarly, Kelly et al. (2003) found that mean annual Spotted Owl occupancy declined after Barred Owls were detected within 0.80 km of territory centers in Oregon and Washington, as compared to Barred Owl-absent territories. Pearson and Livezey (2003) noted about 20% of 129 Spotted Owl sites surveyed from 1996-2001 were unoccupied by Spotted Owls by 2001 in southwestern Washington; they determined there were significantly more Barred Owl site-centers in unoccupied than occupied Spotted Owl home range circles. Kroll et al. (2010) documented lower occupancy probabilities when Barred Owls were presences throughout their study sites in eastern Washington.
Figure 2. Estimates of mean annual site occupancy generated across all Northern Spotted Owl territories in southern Oregon from 1991 to 2006. From Dugger et al. (2011).
Spotted Owl extinction probabilities increased when Barred Owls were present in all six studies for which they were estimated (Dugger et al. 2011, 2016; Kroll et al. 2010; Olson et al. 2005; Sovern et al. 2014, Yackulic et al. 2014). Dugger et al. (2016) found higher extinction rates for the Northern Spotted Owl in all study areas in its range when Barred Owls were present. Olson et al. (2005) noted that the increase in local extinction probabilities affected occupancy probabilities, leading to occupancy rate declines of up to 15% for Spotted Owls. When the two species were modeled together, Yackulic et al. (2014) found that extinction probabilities increased for each species when the other was present, suggesting a strong role of competition in occupancy dynamics.
Simulations have yielded concordant results. For example. Yackulic et al. (2014) found a strong role of intraspecific competition in structuring occupancy dynamics in their modelling. Their simulations suggested competition has a much more substantial impact on the equilibrium occupancy values of Spotted Owls than that of Barred Owls. These simulations also suggest that competition at the patch scale led to increased rates of local extinction for Spotted Owls and, though this would probably not directly drive competitive exclusion from territories, it would result in reduced equilibrium occupancies. Dugger et al. (2011) noted that the strong negative effect of the Barred Owl on occupancy dynamics of Northern Spotted Owls provided evidence of interference competition. By contrast, Bailey et al. (2009) modeled the co-occurrence of Spotted and Barred owls using a two-species occupancy model and found no evidence that Barred Owls excluded Spotted Owls from territories. However, the two species co-occurred less often than expected. Multiple field studies coupled with simulations demonstrate that interspecific competition occurs between Barred and Spotted Owls which has lead to reduced colonization coupled with heightened extinction probabilities of Spotted Owls.
The effect of Barred Owls on Northern Spotted Owl fecundity and recruitment has been mixed. Three of six studies showed no effect of the presence of Barred Owls on Spotted Owl fecundity (measured as number of young fledged per female per year; Anthony et al. 2006, Dugger et al. 2016, Iverson 2004). The other three studies showed a negative, though not always strong, effect (Glenn et al. 2010, 2011b; Olson et al. 2004). Glenn et al. (2010) found that Spotted Owl recruitment rates were diminished in the presence of Barred Owls in four of six study areas. This mixed result may point to other factors influencing fecundity, for instance Wiens (2012) found that the number of young fledged by Spotted Owls increased linearly with increasing distance from the nearest Barred Owl nest or territory center. When Spotted Owl young fledged, Barred Owl detections had no direct relationship with natal or settling locations or dispersal distances (Hollenbeck et al. 2018). They also found that net dispersal distances varied by ecoregions (Washington Coast and Cascades, Washington Eastern Cascades, Oregon Coast Range, Oregon and California Cascades, and Oregon and California Klamath) but declined similarly in all ecoregions over time (~1 km/yr).
Spotted Owl survival was mostly negatively affected by the presence of Barred Owls. Of six studies that analyzed apparent or actual survival, two showed definite negative effects of Barred Owl presence on Northern Spotted Owl survival estimates (Dugger et al. 2016, Glenn et al. 2010), three showed weak negative effects (Anthony et al. 2006, Glenn et al. 2011a, Wiens 2012), and one showed no effect (Schilling et al. 2013).
Spotted and Barred owl diets overlap to some degree in all regions. Hamer et al. (2001) and Wiens et al. (2014) used the Pianka Index to estimate a dietary overlap of 76% in the western Cascades Range in Washington and 44.6% in the central Coast Range in Oregon, respectively. In both studies, Spotted Owl diets were heavily dependent on mammals, which were about 95% of the prey items consumed, especially Northern Flying Squirrels (Glaucomys sabrinus); and while more than half of Barred Owl diets depended on mammals including the Northern Flying Squirrel, they also included a greater percentage of smaller mammals like shrews (Sorex and Neurotrichus spp.) and moles (Scapanus spp.), as well as birds, amphibians, and arthropods (Table 3).
Table 3. Select Barred Owl and Spotted Owl prey items found in regurgitated pellets from three studies in Washington and Oregon. Percent of total number of items identified in owl pellets are listed, with number of pellets sampled in parentheses. No sample size was reported from Hamer et al. (2001).
Barred Owl diets also varies between regions, demonstrating a broad range in prey. Graham (2012) examined Barred Owl pellets from three study areas: Olympic National Forest and the eastern Cascades in Washington, and the Central Coast Range in Oregon. He found that their diet was considerably different between the two western mountain range study areas and the eastern Cascade Range, which is considerably drier and hotter (Table 3). In the eastern mountains, Barred Owls did not depend as much on mammals which was only 26.5% of their diet, but instead relied more on arthropods (47%), as well as amphibians and reptiles (11.1%). Livezey et al. (2008) found that Barred Owls will also eat soft-bodied prey such as earthworms and slugs, which would not be easily detected in regurgitated pellets, upon which these studies depended.
Barred Owls are known to be predators of other owl species, including Strix species on rare occasions (Graham 2012, Wiens 2012). Graham found the remains of two unidentified Strix species near Barred Owl roost trees or nests. Wiens found the remains of two Spotted Owls cached beneath fallen logs with wounds consistent with those inflicted by a large raptor. He was unable to rule out they were not killed by Barred Owls, since Great Horned Owls (Bubo virginianus) were also in the vicinity. Leskiw and Gutiérrez (1998) found a freshly dead Spotted Owl that may have been killed by a Barred Owl. The Barred Owl that flew in next to the authors in response to a Spotted Owl call had feathers similar to those of a Spotted Owl in its talons.
When Spotted and Barred owls interact, the Barred Owl most frequently assumes the dominant role (Van Lanen et al. 2011). Van Lanen et al. conducted experiments using playback tapes and mounted owl taxidermy mounts for both species. They found that male Barred Owls gave more aggressive calls and were more likely to attack the Spotted Owl mount, while male Spotted Owls were less likely to give aggressive calls or attack the Barred Owl mount. Gutiérrez et al. (2004) reported instances where Barred Owls have attacked Spotted Owls, as well as surveyors imitating Spotted Owl calls. Wiens (2012) reported regular interspecific territorial interactions between newly colonizing Barred Owls within the breeding home ranges of Spotted Owls. Interactions included agitated vocalizations by both species near nest sites and Barred Owls chasing Spotted Owls out of shared core-use areas (but not the opposite). In California, Jennings et al. (2011) reported a Barred Owl chasing a female Spotted Owl.
Conversely, there are few observations of Spotted Owl aggressions towards Barred Owls, with the exception of a few reports of nesting Spotted Owls defending a nest or a family group (Gutiérrez et al. 2004) and Jennings et al.’s (2011) report of a pair of Spotted Owls charging and diving at a Barred Owl and an “aerial clash” between a Spotted and Barred Owl (though they did not report how that interaction started).
Crozier et al. (2006) reported that both California and Northern Spotted Owls responded less frequently to Spotted Owl calls after exposure to Barred Owl calls and Northern Spotted Owls responded less frequently in areas having higher numbers of Barred Owls. Other researchers have also noted that the presence of Barred Owls adversely affected Spotted Owl detectability (Bailey et al. 2009, Crozier et al. 2006, Kroll et al. 2010, Olson et al. 2005, Sovern et al. 2014). Olson et al. (2005) found the presence of Barred Owls was important for modeling detectability among years in all occupancy analyses, having a negative effect on Spotted Owl detectability. However, Gutiérrez et al. (2004) noted that even if this were the case, there does not appear to be a decline in Spotted Owl recapture rates, which he stated would be expected if Barred Owls were causing significant behavioral interference.
In consideration of the potential for modified Spotted Owl behavior in the presence of Barred Owls, Wasser et al. (2012) demonstrated an alternate method for surveying both owl species. Their method uses trained “detection dogs” to search out accumulated owl pellets under roost sites, then analyzing the mitochondrial DNA in the pellets to confirm species. The researchers found that detection probabilities using this method were significantly higher for both species than with standard vocalization surveys.
Hybridization between Spotted and Barred Owls
The first Spotted Owl x Barred Owl hybrids were reported in 1987 (Kelly 2001, Kelly and Forsman 2004). Since then, at least 50 hybrids have been reported in the Northern Spotted Owl’s range (Hamer et al 1994, Mazur and James 2000, Pearson and Livezey 2003, Seamans et al. 2004). Kelly and Forsman (2004) reported the very low rate of interspecific matings, thereby suggesting that the rate of hybridization will likely not be a serious threat to the Spotted Owl. However, they also stated, it is possible that hybridization is more common than reported since hybrid backcrosses are hard to identify in all cases. Funk et al. (2007) found that of 12 owls identified as hybrids by plumage in the field, five (almost half) were either Barred (3) or Spotted (2) owls by genetic testing. Hanna et al. (2018) suggested that if hybridization does become more frequent, genetic swamping of the Northern Spotted Owl may occur leading to genetic introgression, which could reduce fitness; however he found no evidence of introgression. Forensic genetic investigation shows that the two species show extensive evolutionary divergence, and that the hybrids are primarily crosses between male Spotted Owls and female Barred Owls (Haig et al. 2004). This makes sense when considering that male owls often present females with food during courtship: male Barred Owls presenting non-mammalian prey to female Spotted Owls would not result in a successful courtship.
As a recent but closely-related invader to the west coast, Barred Owls have the potential to bring novel, harmful pathogens and parasites from east coast populations which could be transmitted to Spotted Owls. As such, Lewicki et al. (2015) examined the Haemoproteus blood parasite assemblages of Barred Owls in both their native and invasive ranges and Northern Spotted Owls. They found that Northern Spotted Owls had a slightly lower prevalence of Haemoproteus infection than both populations of Barred Owls, but mean infection intensity was almost 100 times greater than that of western Barred Owls. They noted their results suggested that Haemoproteus in Spotted Owls are not solely influenced by Barred Owls. They did not directly evaluate if and to what extent parasite infection may have influenced fitness but noted that parasites can become pathogenic with additional stressors, such as competition with Barred Owls.
The future of Spotted Owl and Barred Owl populations
Gutiérrez et al. (2004) outlined and discussed the uncertainty of the ultimate outcome of the invasion of the Barred Owl on the future of these two species. They listed nine potential futures for the Northern Spotted Owl, listed in order of their outcome from most serious to least serious effect:
1. Barred Owls will replace the Northern Spotted Owl throughout its range (behavioral and competitive dominance hypothesis).
2. Barred Owls will replace the Northern Spotted Owl in the northern, more mesic areas of its range (moisture-dependent hypothesis).
3. Barred Owls will replace Northern Spotted Owls over much of its range, but the Spotted Owl could persist in some areas with management intervention (management hypothesis).
4. Barred Owls will replace Northern Spotted Owls over much of its range, but the Spotted Owl will persist in refugia (refugia hypothesis).
5. Barred Owls will replace Northern Spotted Owls in the northern part of its range but the Spotted Owl will maintain a competitive advantage in habitats where its prey is abundant and diverse (specialist vs. generalist hypothesis).
6. Barred Owls will replace Spotted Owls only where weather and habitat change have placed Spotted Owls at a competitive disadvantage (synergistic effects hypothesis).
7. Barred Owls will replace Northern Spotted Owls in some habitats but not in others (habitat hypothesis based on structural elements of forest, which confer a maneuverability advantage to the smaller Spotted Owl).
8. Barred Owls and Spotted Owls will compete, with the outcome being an equilibrium favoring Barred Owls over Spotted Owls in most but not all of the present NSO habitat range (interference competition hypothesis).
9. Barred Owls will increase to a peak number, then decline or stabilize at a lower density, which will permit the continuation of Spotted Owls (dynamics hypothesis). cellspacing="0" width="100%"
A two-day workshop in 2005 was convened to discuss approaches to combat the threat of the Barred Owl invasion to Northern Spotted Owls (Buchanan et al. 2007). Suggestions ranged from no action, to removing Barred Owls using lethal methods. The final recommendation was that emphasis should be on removal experiments and on intensive field studies on aspects of Barred Owl life history and interspecific interactions. Removal of Barred Owls has also been suggested by others (Courtney and Franklin 2004, Gutiérrez et al. 2007), noting that it would have the added benefit of ascertaining the effect of Barred Owls on Spotted Owls.
Diller et al. (2014) began removing Barred Owls on Green Diamond Resources lands in northern California in 2009 as a pilot study to develop methods and study design and to evaluate the feasibility of a removal program. They determined that it was a relatively quick, effective, and low-cost program. Following removal, Spotted Owl occupancy made a slow recovery in areas where Barred Owls were removed compared to untreated areas where occupancy declined, while Barred Owl occupancy declined (Diller et al. 2016). Spotted Owl fecundity rates did not change in the treated areas, however the greater number of pairs in treated areas resulted in greater overall population productivity; there was also an increase in estimated survival and population change, and a decrease in extinction rate. Barred Owl removals continued through 2013, resulting in an increased rate of population change as well as realized increased population change (Dugger et al. 2016). In 2015, Wiens et al. (2017) began a 5-year Barred Owl removal study in established study areas in Washington and Oregon. After removal, they noted a decline in the probability of use by territorial pairs of Barred Owls on both the control and treatment areas between 2015 and 2016, which they felt might indicate a decline not directly attributable to experimental removal activities.
The question remains if such an approach would actually be successful and feasible throughout the Northern Spotted Owl range. Bodine and Capaldi (2017) used mathematical modelling to determine if removing Barred Owls could ultimately save the Northern Spotted Owl population in the Oregon Coast Range. Their models showed that Barred and Northern Spotted owls could not coexist in the long run, suggesting that Barred Owls would need to be eliminated. They determined that by 2030, less than 5% of historical Spotted Owl sites would contain Spotted Owls. However, removal of Barred Owls to eliminate them would require clearing about 50% of the Barred Owl sites per year to eliminate them in over 100 years. To eliminate them in 10 years, it would take removing over 90% of Barred Owl sites yearly. Perlman (2017) examined the effects of various Barred Owl removal rates in four existing demographic Northern Spotted Owl study areas using a two-species individual-based model. She found that viable, long-term recovery was observed for only two of the four areas, and was dependent on high removal intensities. No model showed a return to pre-invasion population numbers. However, her models agree with Bodine and Capaldi, in that with no removal action taken, the Northern Spotted Owl population would go extinct within 100 years. Baumbusch (2016) investigated the effect of size of patch on removals with modelling and found that large contiguous removal areas maintained a lower Barred Owl occupancy and required removal of fewer owls compared to small fragmented removal areas covering the same acreage.
By contrast, Yackulic et al. (2014) noted that complete eradication of Barred Owls is unlikely, but that it might be more feasible to maintain their occupancy at a low level (0.2), benefitting Spotted Owls while decreasing cost. They suggested that initial removal effort be focused on patches with high quality habitat, where Spotted Owls have typically had high reproductive rates or are currently or have recently been observed to be occupied by Spotted Owls. Holm et al. (2016) suggested that removal efforts would be most effective in areas with low Barred Owl populations and where it would be defensible against Barred Owl colonization or have high-quality Spotted Owl habitat. Yackulic et al. (2012) noted that understanding the influence of regional occupancy on local colonization and extinction rates of Barred Owls will be important to predict the impacts of Barred Owl removal as a management tool.
Diller et al. (2014) found the cost for such removals not to be overly burdensome, costing up to $150/owl in direct costs. On the other hand, Livezey (2010) estimated cost to be about $1 million annually for the entire Northern Spotted Owl range, including additional indirect costs, breaking down to about $700/owl in the first year and $2800/owl in subsequent years. Ultimately, there are also questions about interfering with a natural process of competition and how long any removal program could last, since there would still likely be immigration from the east, if not from the much closer Barred Owl populations in Canada to the north (Bodine and Capaldi 2017, Cornwall 2014). Public acceptance of such methods may also be difficult to overcome (Lute and Attari 2017). Other authors noted that habitat management may aid the Spotted Owl or that a natural balance between Barred and Spotted Owls has or can be achieved at least in certain areas, allowing for coexistence between the species (Buchanan et al. 2007, Pearson and Livezey 2007).
Important information gaps
• Effects of disturbance on Barred Owls. The influence of fuel treatments and thinning on Spotted Owl occupancy has been repeatedly explored. Determining how variation in such treatments affects Barred Owl colonization, and their subsequent effects on Spotted Owls, represents an important line of inquiry.
• Interactive effects of Barred Owls and fire on Spotted Owls. Barred Owls can use a wider range of habitats and food resources than Spotted Owls. Such behavioral flexibility may make Barred Owls relatively more successful in burned landscapes than Spotted Owls. To our knowledge this conjecture remains unverified.
• Recolonization rates of Barred Owls after their removal. The positive response of Spotted Owls to Barred Owl removal has been encouraging. However, the rate at which Barred Owl recovery occurs post-removal is unknown and would greatly inform planning for future removal efforts. cellspacing="0" width="100%"
Key Messages for Managers
• Barred Owls reduce Spotted Owl occupancy, survival, productivity, fecundity, and population growth.
• Barred Owls exhibit a wide dietary breadth and are likely influencing food webs in unforeseen ways.
• Be cautious in assuming that recently arrived Barred Owls will remain in low numbers in perpetuity. Barred Owls have repeatedly shown that they occur in low numbers prior to dramatically increasing in abundance.
• Barred Owls appear to suppress detection probabilities of Spotted Owls. Consider alternative techniques, such as scat sniffing dogs, or increasing survey effort to detect Spotted Owls.
• Experimental removal of Barred Owls has proven effective in improving Spotted Owl population growth.
• Maximize the value of Barred Owl removals by focusing on those habitats most valuable for Spotted Owls with the least number of Barred Owls, prior to removals in poor habitat with high numbers of Barred Owls.
Annotated Bibliography of Cited References