Introduction

Pyrodiversity, the mean spatial variability in wildfire effects, results in complex post-fire vegetation mosaics that are associated with high levels of biodiversity. Large fires that produce a variety of severities (i.e., mixed-severity fires) in ponderosa pine (Pinus ponderosa Laws) and mixed-conifer forests of the western USA are increasingly recognized for their importance in generating pyrodiverse landscapes (e.g., Perry et al. 2011, Williams and Baker 2012, Odion et al. 2014, Marcoux et al. 2015). Top-down processes such as extreme fire weather, regional climate (which influences fuel moisture and ignitions), and bottom-up processes such as topographic relief, vegetation, and disturbance history govern the distribution and size of fire patches in mixed-severity fires (Perry et al. 2011, Dunn and Bailey 2016). Regional drought, high winds and temperatures, and other factors (e.g., surface fuel loading, crown base height, and crown bulk density; Cruz and Alexander 2010) drive crown fire behavior in these systems, producing small and large patches of high tree mortality within a predominantly surface-fire matrix of mostly surviving trees. Mixed-severity fires therefore generate complex stand structures and landscape heterogeneity—characteristics not typically produced by low-severity fire (Table 1). Low-severity fire, while also important ecologically, is preferred by many managers due to lower risks to economic values. Here, we focus on mixed-severity fires because they have received less attention by managers, but they result in pyrodiverse landscapes (DellaSala and Hanson 2015). We demonstrate how ecosystem integrity can be met by managing for pyrodiverse landscapes mediated by mixed-severity fires in the biodiverse region of the Sierra Nevada, California, USA.

Table 1 Pyrodiversity attributes produced by mixed-severity fires associated with high levels of biodiversity and ecosystem functions.

Although it is the subject of ongoing research and debate (Odion et al. 2016), it has been suggested that mixed conifer, ponderosa pine, and Jeffrey pine (Pinus jeffreyi Grev. & Balf.) forests in this region historically experienced a mix of fire severities, including areas of high overstory tree mortality (DellaSala et al. 2014, Stevens et al. 2016). There is considerable variability in reported proportions and sizes of high-severity fire patches, with the greatest differences found in relatively smaller study areas or studies in which shorter time periods were analyzed (Table 2). High-severity patches commonly ranged from 0.4 ha to >50 ha, but the historical frequency of patches >1000 ha is still debated (e.g., Baker 2014, Stevens et al. 2016). While uncertainty remains on some issues, there is general agreement that most forests of the Sierra Nevada currently have less high-severity fire, in terms of annual or decadal area burned, than they did historically, prior to fire suppression (Mallek et al. 2013, Odion et al. 2014, Baker 2015). Additionally, drier low-elevation pine forests burned most frequently at low to moderate severity (Stephens et al. 2015), but those fires also contained variably sized high-severity patches (Leiberg 1902, Baker 2014, Hanson and Odion 2016 a, b). Douglas-fir (Pseudotsuga menziesii Mirbel) (Odion et al. 2014) and Sierra lodgepole pine (Pinus contorta var. murrayana Grev. & Balf.) forests experienced mixed-severity fires as well (Caprio 2008).

Table 2 Historical fire severity proportions and maximum high-severity fire patch sizes in mixed-conifer and ponderosa pine forests, Sierra Nevada management region.

Tree mortality is also an important component of mixed-severity fire effects characterized mostly by low-mortality levels (0 % to 20 % tree basal area), highly variable moderate-mortality levels (20 % to 70 %), and high-mortality levels (>70 % tree mortality) (Perry et al. 2011; Figure 1). Agee (2005) noted that mixed-severity fires are not merely an intermediate state between low and high severity but, rather, are a unique type of disturbance that warrants careful study by ecologists.

Figure 1
figure 1

(A) Landscape view of mixed-severity fire effects in the Rim Fire 1 year post fire. The spatial pattern of fire severity patches and patch sizes results in a pyrodiverse landscape that provides habitat for wildlife across a post-fire vegetation gradient of low or unburned vegetation patches to severely burned vegetation patches. (B) Close-up of large patch of complex early seral forest created by high-severity fire in juxtaposition with abundant and varied “biological legacy” trees (complex structures, such as snags, logs, and shrubs that survive fire). Photos by C. Hanson.

While there are winners and losers in the immediate aftermath of any disturbance event, the net effect of mixed-severity fire is that it provides a mosaic of habitat for a broad suite of species. For instance, songbirds have high levels of species richness and abundance in post-fire vegetation at mid elevations (Fontaine et al. 2009, Tingley et al. 2016). Black-backed woodpeckers (Picoides arcticus [Swainson, 1832]), mountain bluebirds (Sialia currucoides [Bechstein, 1798]), tree swallows (Tachycineta bicolor [Vieillot, 1808]), and numerous shrub-nesting birds preferentially use recently burned forests in the Sierra Nevada and other regions, presumably due to increased shrub cover and presence of snags (Fontaine et al. 2009, DellaSala et al. 2014, Hutto et al. 2015, Tingley et al. 2016). California spotted owls (Strix occidentalis occidentalis [Xantus de Vesey, 1860]) and olive-sided flycatchers (Contopus cooperi [Nuttall, 1831]) forage in severely burned patches where prey are abundant, and nest in unburned to moderately burned portions of the same fire mosaic (Bond et al. 2009, 2016; Hutto et al. 2015; Comfort et al. 2016). Bats make use of high snag densities (Buchalski et al. 2013) and fire-recruiting plants are associated with severely burned patches (Donato et al. 2009). Even mature-forest carnivores such as the Pacific fisher (Pekania pennanti [Erxleben, 1777]) actively forage in severely burned patches (Hanson 2015).

The high-severity patches within the mixed-severity mosaic provide a unique pulse of biological legacies—complex structures such as snags, downed logs, and native shrub patches from seed that survive fire and that are important in connecting seral stages through time (Franklin et al. 2000, Fontaine et al. 2009, Donato et al. 2012, DellaSala et al. 2014). The economic value of large dead and live trees within these patches means that commercial trees are most often targeted for harvest soon after fire. In addition, nursery-grown young trees are planted soon after fire and, to promote the crop of young trees, herbicides are often sprayed to kill competing vegetation (Lindenmayer et al. 2008, 2017). Logging slash from post-fire logging may contribute to subsequent fire behavior (Donato et al. 2006, Thompson et al. 2007), as can the fuel array of densely planted even-aged trees (Odion et al. 2004).

On public lands, current fire policy promotes thinning over large landscapes (e.g., USDA Forest Service 2002, US Congress 2003, USDA Forest Service 2009, US Congress 2015), which is costly (Schoennagel and Nelson 2011), infeasible over large areas (Calkin et al. 2013, North et al. 2015 a, Parks et al. 2015), and largely ineffective under extreme fire weather conditions (Lydersen et al. 2014, Cary et al. 2016). For instance, from 2001 to 2008, over 11 million hectares were thinned on national forests (mostly in the western USA) at a cost of more than $6 billion (Schoennagel and Nelson 2011). Mechanical vegetation treatments can cost over $3700 per hectare for each round of thinning (Kline 2004), which would need to be repeated at least every 15 to 20 years to keep flammable vegetation at low levels. Additionally, from 1985 to 2015, suppression costs were more than $25 billion to fight approximately 2 million fires on over 83 million hectares, mostly spent by the Forest Service (Ingalsbee and Raja 2015).

Thus, we concur with others that active management approaches could include more natural fire ignitions (Calkin 2013, Meyer 2015, North et al. 2015 b) or resource objective wildfires (Meyer 2015) in which fire is put back on the landscape to hasten the process of forest restoration (Moritz et al. 2014, Moritz and Knowles 2016). This would also help to meet fire and fuels objectives and allow managers to better accommodate mixed-severity fire effects for ecosystem integrity (Meyer 2015, Dunn and Bailey 2016). We suggest that an ecosystem integrity approach is not inconsistent with current active fuel management on federal lands and may be a cost-effective way to achieve biodiversity goals (North et al. 2015 b), while reducing some of the conflicts associated with extensive fuels-focused approaches—particularly impacts to imperiled species and at-risk ecosystems. We use the definition of ecosystem integrity common in the literature (e.g., Pimentel et al. 2000), also adopted by the USDA Forest Service (2012), as the ability of an ecological system to support and maintain a community of organisms that has a species composition, diversity, and functional organization comparable to those of natural habitats within a region.

Our focus is the Sierra Nevada region because of national attention given to so many recent fires therein. We include an example of a fire-adapted species (black-backed woodpecker) that uses high-severity patches, and an imperiled species (California spotted owl) known to decline within intensively managed post-fire landscapes. The Sierra Nevada is one of the most diverse temperate conifer forest regions on Earth and has exceptional levels of plant endemism (Ricketts et al. 1999). Approximately half of California’s 7000 vascular plant species occur in this region, with 400 considered endemic and 200 rare. High levels of vertebrate richness and endemism also occur. Species composition varies across north-south, east-west, and elevational gradients, resulting in high levels of beta diversity.

Importantly, the 2012 forest planning rule (USDA Forest Service 2012) includes specific provisions for managing public resources to maintain or restore: (1) structure, function, composition, and landscape connectivity; (2) ecological conditions for recovery of imperiled and focal species; and (3) rare and unique habitat types (USDA Forest Service 2012). The National Cohesive Wildland Fire Management Strategy (USDI and USDA 2014) and Sierra national parks (e.g., Yosemite, Sequoia and Kings Canyon) also include multi-faceted approaches that promote greater wildfire ignitions. Though national forest lands compose most of the forested area in California, and are thus our focus herein, significant areas of federal forest in California are managed by the National Park Service (NPS), and a state agency, California Department of Forestry and Fire Protection (CAL FIRE), responsible for decisions and operations pertaining to fire suppression on private and state lands. NPS, like the Forest Service, is required to protect species listed under the Endangered Species Act (ESA), and CAL FIRE is subject to the California state ESA. Thus, our approach to wildfire management can be applied to these agencies and land ownerships regarding decisions about fire suppression and forest management that might impact imperiled or ESA-listed species associated with post-fire landscapes.

Study Area

The Sierra Nevada management region is a 750 km long, north-south oriented mountain range in California composed of granitic rock, and distributed across three ecoregions: Sierra Nevada proper; portions of the Modoc Plateau; and the eastern portion of the southern Cascades (Bailey 1995; Figure 2). The regional climate is mediterranean with cool, wet winters, and warm, dry summers; precipitation generally decreases west to east and north to south (Millar1996).

Figure 2
figure 2

Sierra Nevada study region showing national forests, national parks, and inventoried roadless areas.

There are 11 national forests totaling about 4.6 million hectares: Modoc, Lassen, Plumas, Tahoe, Eldorado, Stanislaus, Sequoia, Sierra, Inyo, Humboldt-Toiyabe (western portion), and Tahoe Lake Basin Management Unit. Forest planning is governed by the Sierra Nevada Framework (USDA Forest Service 2004), but the Forest Service is currently revising its forest plans for the Inyo, Sequoia, and Sierra national forests as “early adopters” (i.e., first national forests to test the planning rule) of the 2012 forest-planning rule (USDA Forest Service 2012). Three national parks—Lassen, Sequoia and Kings Canyon, and Yosemite— and several large wilderness and inventoried roadless areas >2000 ha also occur in the region.

Coarse-Filter and Fine-Filter Approaches to Ecosystem Integrity in Mixed-Severity Systems

Managers wishing to maintain ecosystem integrity via naturally ignited fires can do so using a combination of coarse- and fine-filter conservation approaches (Noon et al. 2003, USDA Forest Service 2012). Coarse filters invariably include relatively few indicators associated with the larger ecosystem of interest (e.g., major vegetation types or, in this case, different categories of burn severity). Their presence is meant to indicate that essential components of the whole system are intact, and they operate at broad spatial scales such as those associated with large fires (hundreds of square kilometers). Coarse filters are typically used to guide reserve design based on fundamental principles of conservation biology, including spatially redundant reserve complexes representative of the major forest types and fire severities interconnected across large landscapes. To achieve a pyrodiverse landscape, perhaps the best coarse filter would include high-severity fire patches interspersed with fire refugia (unburned areas) and low- to moderate-severity patches.

Fine-filter considerations complement coarse filters by adding site-specific or habitat elements associated with focal species, guilds, or other species groupings (USDA Forest Service 2012). Application of this kind of filter allows managers to evaluate whether habitat and special conservation needs are met through a given management plan, and ground-truth the utility of burn severity maps by linking mapped fire severities to habitat needs of target species. In addition, the approach allows managers to meet national forest planning requirements to monitor and evaluate a small suite of focal species selected to assess the degree to which ecological conditions are supporting the diversity of plant and animal communities within a given planning area (USDA Forest Service 2012). Focal species can, therefore, be used to monitor the integrity of the larger system to which they belong, and researchers (e.g., Seavy and Alexander 2014, Stephens et al. 2015, Siegel et al. 2016) have suggested using patterns of plant and animal distributions as a passive management strategy to accommodate mixed-severity systems. The Forest Service also now considers species of conservation concern as “a species, other than federally recognized threatened, endangered, proposed, or candidate species, that is known to occur in the plan area and for which the regional forester has determined that the best available scientific information indicates substantial concern about the species’ capability to persist over the long-term in the plan area” (36 CFR 219.9(c); https://www.fs.usda.gov/Internet/FSE_DOCUMENTS/stelprdb5359595.pdf, accessed 12 May 2017). The agency is required to maintain suitable habitat for these species to ensure viable populations are present in the planning area (USDA Forest Service 2012).

Comprehensive Wildland Fire Management

We recognize that land managers face many constraints (legal and social) and often competing regulatory and management objectives that limit wildfire management options. However, the Planning Rule and the National Cohesive Wildland Fire Management Strategy (USDI and USDA 2014) offer opportunities to put more fire back on the landscape whether through prescribed burning or managed wildfires. We provide some general concepts that managers might apply with pyrodiversity outcomes realized through mixed-severity fires that meet ecosystem integrity objectives.

Integrating Wildland Fire and Targeted Fire Suppression (Coarse Filter)

Mixed-severity fire effects for ecosystem benefits can be integrated with targeted suppression and fire-risk reduction efforts near towns using this coarse-filter approach. While we acknowledge that there was concern about the size and severity of the 2013 Rim Fire (Lydersen et al. 2014), the largest fire in recent Sierra Nevada history, we note that even this fire produced mostly low- to moderate-severity effects (i.e., ∼20 % of the burn was high severity based on Monitoring Trends in Burn Severity [MTBS]; http://mtbs.gov/MTBS_Uploads/data/2013/maps/ca3785712008620130817_map.pdf, accessed 23 April 2017), and a wide range of high-severity patch sizes, which contributed to significant heterogeneity at landscape scales. Thus, we concur with others (e.g., Moritz et al. 2014, Ingalsbee and Raja 2015, Dunn and Bailey 2016, Moritz and Knowles 2016, Schoennagel et al. 2017) that suppression could be focused narrowly to lands surrounding towns and used in combination with defensible space management nearest homes (Cohen 2000, 2004) so that more wildland fires can burn safely in the backcountry.

Notably, one way to safely modify fire suppression activity would be to restrict large fire crews and heavy equipment to protect homes and communities within the Wildland Urban Interface (WUI). The WUI is usually considered to extend to ∼2 km from an at-risk community (US Congress 2003, USDA Forest Service 2004), even though most vegetation treatments are conducted farther from communities (Schoennagel et al. 2009). Beyond the WUI, point protection strategies would be used to keep fire away from isolated structures and infrastructures like cabins, communication towers, bridges, or other human assets that could be destroyed by fire. Relatively small, mobile fire crews would also use minimum impact suppression tactics (i.e., Minimum Impact Suppression Tactics [MIST]; https://www.nifc.gov/PUBLICATIONS/redbook/2003/AppendixU.pdf, accessed 12 May 2107) in backcountry areas, primarily monitoring fire spread but, when necessary, actively managing it (rather than containing and controlling wildfire as in traditional full-suppression strategies) by steering fire away from threatened social assets (Donovan and Brown 2005, 2008; Ingalsbee and Raja 2015). In municipal watersheds where fire management plans may want to avoid high-severity fires burning near water sources, more fires could be allowed to burn during moderate weather conditions. Wildfire management should be a useful tool for managing fuel loads in municipal watersheds where the use of chemicals or heavy equipment for either thinning or suppression would cause unacceptable impacts to water quality and soils. MIST could also be employed where fires in wilderness and roadless complexes, national parks, and even in roaded areas many kilometers from the nearest town pose low risk to residential areas. In sum, this approach would shift wildfire operations from limiting fire spread, size, or duration in backcountry areas to working with fire for ecosystem benefits while still effectively providing for community wildfire protection.

Sierra Nevada national forests and parks are large enough to accommodate most large fires over thousands or even tens of thousands of hectares (Appendix 1). For instance, many (>50 %) of the largest forest fires from 1984 to 2014 were primarily contained within an individual national forest or national park boundary. In general, federal lands offer unique opportunities in which the maintenance of pyrodiversity for biodiversity could be emphasized in large protected areas (wilderness and roadless area complexes; Appendix 2). Coordination among agencies with similar objectives may allow for more naturally ignited fires over mixed ownerships having similar objectives (e.g., wilderness or roadless areas, other remote forests, conservation areas juxtaposed with parks) using an all-lands approach. If reserves were too small to accommodate large fires or patches of different fire severities, then complexes of multiple reserves widely distributed across a region in redundant locations would collectively help maintain the full complement of post-fire stages using the coarse-filter approach.

In the Sierra Nevada, the draft revised forest plans for the three early-adopter national forests in the southern portion of the range have included a fire-management-zoning approach similar to what we suggest here, allowing more naturally ignited fire in remote areas and suppressing fires close to communities (USDA Forest Service 2016). However, the focus in the draft plans remains on mechanical thinning and post-fire logging (USDA Forest Service 2016). We submit that an approach that allows more natural fire ignitions is advisable and warranted from the standpoint of both ecosystem integrity and public safety, as discussed herein.

Focal Species and Species of Conservation Concern (Fine Filter)

By way of example, we consider two species that could be used to monitor mixed-severity effects. The black-backed woodpecker would be an ideal focal species given its very close association with high-severity fire patches, as would the California spotted owl, a species of conservation concern. Both species are complementary to mixed-severity fire management, given that the woodpecker is mainly associated with the high-severity component, and spotted owls use a broad gradient of fire severity patches. Moreover, while there is some overlap in geographic ranges, spotted owls generally occupy low- to mid-montane forests, while the black-backed woodpecker lives in mid- to high-elevation mixed-conifer forests up to subalpine forests.

Black-backed woodpecker as focal species of high-severity fire patches. In the Sierra Nevada, black-backed woodpeckers occur across mid- to upper-montane and subalpine conifer forests from ∼1200 m to 2800 m, depending on latitude. While still uncommon even in burned areas, the greatest concentrations occur in severely burned, mixed-conifer and upper montane forests with high basal area of snags (Hanson and North 2008, Saracco et al. 2011) where wood-boring beetle larvae are abundant (Saab et al. 2007). Burned areas also typically harbor high densities of medium to large dead trees >30 cm dbh (Cahall and Hayes 2009, Saab et al. 2009, Tingley et al. 2014). Black-backed woodpeckers also occur (albeit much more rarely) in dense, mature unburned forests (Bonnot et al. 2009, Fogg et al. 2014) where they have relatively larger home ranges, presumably reflecting conditions that are less than optimal (Tingley et al. 2014). Nevertheless, unburned forests with high levels of dead trees from drought and native bark beetles might at least slow the rate of population decline during interludes between severe fires (Rota et al. 2014). Only a small fraction of fires burn suitable woodpecker habitat, due to the narrow convergence of conditions that include recent (generally ≤8 years post-fire) higher-severity fire effects in dense, mature, middle- to high-elevation conifer forest (Casas et al. 2016). Often a single pair of birds uses hundreds of hectares (Dudley and Saab 2007, Tingley et al. 2014).

Black-backed woodpeckers are vulnerable to even partial post-fire logging (Hutto and Gallo 2006, Koivula and Schmiegelow 2007, Saab et al. 2009, Rost et al. 2013). Radio-telemetry studies in the Lassen and Plumas national forests of California showed that home-range sizes were significantly larger in forests in which some post-fire logging occurred, and post-fire logged patches in the Sierra Nevada were avoided (Tingley et al. 2014). For example, even though post-fire logging was proposed for what seems like a minor portion of the King Fire, logging was especially concentrated within the highest quality woodpecker habitat (Figure 3), where a high density of medium to large snags occurred. Notably, on national forests of the Sierra Nevada, post-fire logging decisions have typically authorized removal of 40 % to 60 % of high-severity patches, displacing complex early seral forest with tree plantations (e.g., USDA Forest Service 2014, 2015, 2016). Retention of dead trees in logging units generally averages ∼10 trees per hectare >38 cm dbh (USDA Forest Service 2004). By comparison, to maintain habitat for this focal species, generally hundreds of medium to large snags per hectare (>30 cm dbh to 40 cm dbh, and especially snags >50 cm dbh) are needed (Hanson and North 2008, Saab et al. 2009, Tingley et al. 2014) in patches consistent with home-range size, along with an ample supply of dense, mature or old conifer forest to facilitate conditions for high quality habitat when fires do occur (DellaSala et al. 2014).

Figure 3
figure 3

King Fire logging units on the Eldorado National Forest and black-backed woodpecker nests and sightings. After extensive surveys for black-backed woodpeckers were conducted for the US Forest Service throughout the fire area one year post fire, using playback recordings to detect the birds, all but one of the detections was in a relatively small area of dense, mature mid-montane conifer forest in a very large high-severity fire patch in the northern portion of the fire area (shown above). The Forest Service’s decision authorized post-fire logging of ∼80 % of these locations.

California spotted owl as species of conservation concern. Early studies on habitat associations and reproductive success of spotted owls in the Sierra Nevada were conducted in long-unburned forests, and “non-suitable” owl habitat was typically the result of logging (e.g., Moen and Gutiérrez 1997, Blakesley et al. 2005). Because spotted owls are usually associated with older, dense forests, it was assumed that effects of high-severity wildfires were similar to logging (Weatherspoon et al. 1992). However, recent studies have demonstrated that occupancy (Roberts et al. 2011, Lee et al. 2012, Lee and Bond 2015 a) and reproductive success (Roberts 2008, Lee and Bond 2015 b) were similar or higher in forests burned with a mixture of fire severities compared to long-unburned forests for up to at least 15 years post fire (longer-term studies have not been conducted). Lee and Bond (2015 a) reported higher occupancy rates than any Sierra Nevada study area for historical owl breeding sites one year after the Rim Fire. The amount of high-severity fire within an owl pair’s 120 ha protected activity center, as defined by the Forest Service, had no effect on occupancy, although occupancy by single owls declined slightly as the extent of severe-fire patches increased.

Thus, even though spotted owls are not considered a fire-dependent species, they do persist after mixed-severity fires when both unburned and severely burned patches occur within historical territories (Lee et al. 2012; Lee and Bond 2015a, b). Owls foraged preferentially in high-severity patches within mature forest in the southern Sierra Nevada (Bond et al. 2009) and used high- and moderate-severity patches in the San Bernardino Mountains in proportion to availability (Bond et al. 2016). Notably, structural complexity (including high density of dead trees) is important for spotted owl foraging habitat. Bond et al. (2009) found that dead tree basal area and shrub cover were highest in high-severity fire patches in which owls preferentially foraged. The owls found a rich food source, in the form of small mammal prey, in post-fire habitat (Bond et al. 2016). California spotted owls also selected high-severity patches for foraging more than any other fire severity condition or than long-unburned forests when within 1.5 km of the nest or roost (Figures 4 and 5). Although there are reports of California spotted owls nesting in moderate-severity patches, these raptors mostly nest and roost in long-unburned or lower-severity areas within a burned landscape (Bond et al. 2009), underscoring the importance of the mixed-severity mosaic. In contrast, Jones et al. (2016) found higher rates of territory extirpation and lower rates of colonization of owl sites that experienced >50 % high-severity fire in the King Fire on the Eldorado National Forest, and reported avoidance of high-severity patches for foraging. The circumstances of their study differed greatly from others (Lee and Bond 2015 a, b), presumably due to pre- and post-fire logging within owl territories, as well as extensive high-severity fire in pre-fire clearcuts with young plantations.

Figure 4
figure 4

(A) Estimated foraging locations (obtained in 2006) of seven radio-marked California spotted owls in the 2002 McNally Fire, Sequoia National Forest, Sierra Nevada, USA. Different colored points represent each individual owl’s estimated foraging location. Circles represent foraging ranges: each circle is centered on the nest with its radius extending to the farthest estimated foraging location for each individual owl. White areas are non-suitable for owls (e.g., foothill chaparral vegetation).

Figure 5
figure 5

(A) General location of a California spotted owl nest territory in the 2002 McNally Fire (circle not to scale). Nest site was in a low-severity patch directly adjacent to high-severity patch (severity defined using Miller and Thode 2007). (B) Zoom-in (center snag) of general location of California spotted owl nest tree within McNally Fire burn patch shown in (A). Photos by M. Bond.

Long-term occupancy monitoring without the confounding influence of post-fire logging is especially important to understanding fire effects on spotted owls. Hence, Bond et al. (2009) recommend that, if managers want to maintain spotted owl habitat after fire, they should prohibit post-fire logging and pesticide and herbicide applications within at least 1.5 km of historical spotted owl nest and roost sites. Even larger areas may be needed given that owl breeding-season home ranges can extend upwards of 700 ha (Bond et al. 2016), and some birds expand their range or migrate during the non-breeding season (Bond et al. 2010). Therefore, a reasonable protected area might be within 2.4 km of nest and roost sites, which corresponds to interim spotted owl management guidelines of the Forest Service’s Pacific Southwest Research Station (http://www.fs.usda.gov/Internet/FSE_DOCUMENTS/fseprd504726.pdf).

Restoration of Degraded Forests

Land-use stressors that degrade or impair ecosystem processes are fundamentally at odds with ecosystem integrity approaches (Pimentel et al. 2000, USDA Forest Service 2012). Thus, restoration treatments can be used to reverse the causative agents of ecosystem degradation. One example is to limit human-set fires via: (1) seasonal closure and decommissioning of roads, or convert roads not considered essential in firefighting within the WUI to indefinitely closed; and (2) focused thinning and prescribed burning nearest homes, around campgrounds and other facilities, and along narrowly defined road prisms close to towns to avoid fire spread from anthropogenic ignitions. Managers could also concentrate thinning of small trees (shaded fuel breaks) along with prescribed burning nearest critical evacuation routes for communities with only one means of ingress or egress, redesign traveler stopping points along roads to avoid fire-prone settings, and concentrate visitation in fire-safe locations. Importantly, because tree plantations create unnaturally homogenized forests that lack complex structures, managers could integrate thinning with mixed-intensity prescribed burning, or naturally ignited fires, and create snags and downed logs to introduce structural complexity. Thinning small trees combined with prescribed fire (Kalies and Kent 2016) may reduce fire intensity in densely stocked tree plantations (Odion et al. 2004).

Conclusions

The 2012 Planning Rule provides the Forest Service with new direction for restoring and maintaining integrity and for managing focal species and species of conservation concern that can be integrated with fuels management approaches. The National Cohesive Wildland Fire Management Strategy (USDI and USDA 2014) allows managing wildfire for ecosystem benefits; hence, our findings can be applied to Department of Interior lands as well.

We suggest that managing for ecosystem integrity using both a coarse- and fine-filter approach centered on pyrodiverse fire effects can inform forest management in a biodiversity context. Our approach would have the added benefit of likely reducing suppression costs and some of the negative effects of mechanical vegetation removal over large areas (Dale 2006, Donovan and Brown 2008, Dunn and Bailey 2016). The complementary nature of conservation filters would allow managers to check burn severity maps with habitat associations of focal species to assess management efficacy.

Managers face substantial political and public pressure to suppress fires through the use of aggressive firefighting tactics, but such tactics do little to contain fires under extreme weather conditions (Lydersen et al. 2014, Moritz et al. 2014, Ingalsbee and Raja 2015, Carey et al. 2016). Instead, managers could be encouraged to use prescribed and naturally ignited fires that yield both cost savings and ecosystem benefits. Unfortunately, federal fire suppression budgets are dominated by suppression costs, causing siphoning of funds away from other essential programs (Ingalsbee and Raja 2015). To support managers in using more natural fire ignitions, conditions and certain trigger points could be more clearly defined and integrated with forest planning. This would allow flexibility to use several approaches to managing a fire, even on the same incident. Thus, in theory, a large fire could be managed in one area with general containment strategies that employ MIST (backcountry), while simultaneously in another area (near towns) with direct attack methods.

Accommodating mixed-severity fires for ecosystem benefits pertains to both ends of the fire continuum: large fires with high-severity effects that generate unique biological pulses (e.g., complex structures), and lower-severity systems that may have been homogenized through management and suppression. This suggests an important opportunity for expanding fire management beyond traditional kinds of prescribed burning to include prescriptions that benefit a broader suite of species associated with pyrodiverse landscapes (Moritz et al. 2014, DellaSala and Hanson 2015, Moritz and Knowles 2016). We note the conundrum of natural fire ignitions creating greater smoke emissions that may conflict with air quality objectives. Importantly, the Environmental Protection Agency (2016) recently revised policies to provide special regulatory exemptions and provisions that allow for more managed wildfires.

With proper planning and use of modern smoke management techniques, adverse effects of emissions on public health can be mitigated and fire restoration goals better accommodated. However, smoke emissions must be viewed as an unavoidable trade-off to be weighed against other potentially worse effects from attempted fire exclusion (that will eventually burn in a wildfire) or other chemical and mechanical methods for managing fuel loads that have ecosystem consequences.

There is clearly a need for research on whether natural fire ignitions can primarily provide desired mixed-severity fire effects. We suggest that studies are needed to determine the following.

  1. (1)

    Specific locations and forest types best suited for mixed-severity fire effects, particularly in relation to ecological mechanisms by which pyrodiversity influences biodiversity.

  2. (2)

    Current versus historical sizes and proportions of fire-severity patches and how those might be affected by climate change.

  3. (3)

    Additional species that may be affected by suppression such as declining shrub-nesting birds associated with complex early-seral forests (Hanson 2014).

  4. (4)

    Importance of other disturbance events (e.g., native insect outbreaks, drought) in maintaining ecosystem integrity.

  5. (5)

    Effects of mechanical treatments before and after fire on the integrity and quality of mixed-severity patches including species of conservation concern and focal species.

  6. (6)

    Kinds of education efforts required to implement this type of integrated disturbance ecology approach.

  7. (7)

    Decision-support tools to help managers assess the costs and benefits of natural fire ignitions, along with conditions under which fires should be suppressed for human safety.

We argue that expanding natural fire ignitions for ecosystem benefits in combination with strategic use of defensible space, directed suppression, and active fuels management in appropriate areas provide untapped potential to enhance ecosystem integrity while protecting people and infrastructure with the potential for lower financial costs. Our approach is based on an ecological understanding of the importance of mixed-severity fires (DellaSala and Hanson 2015), and the need to reconsider “catastrophe” biases regarding natural disturbance processes (Lindenmayer et al. 2017).