Overview

In many forested landscapes across western North America, past fires often act as barriers to fire spread for a time and then, as live and dead fuels accumulate, reburn but with much lower severity than surrounding forested areas. In this project, we evaluated the interactions of past burn mosaics within recent large fire events in three study areas to support wildfire management decisions in semi-arid forests. A key inspiration for this project was a map of known fire starts within the 2006 Tripod Complex fire in North Central Washington State. Within the 175,000 acre fire area, over 300 active fire starts were suppressed between 1940 and 2005. In central Idaho, the 320,000-acre East Zone fire had a total of 977 recorded fire starts that were actively suppressed prior to the wildfire event. We conducted this study to better understand the role of past fires as a type of fuel reduction treatment and to also evaluate the long-term consequences of fire suppression.

With funding from the Joint Fire Sciences Program, we evaluated past fire mosaics and their interaction with subsequent large fire events. The study was designed to support the National Wildland Fire Cohesive Strategy and specifically address its goals to increase landscapes resiliency to fire and to improve firefighter safety.

Our project had two main tasks:

• In Task 1, we used burn severity analysis to analyze the effect of past fires on subsequent fire spread and severity.
• In Task 2, we used fire simulation modeling to ask how different wildfire management strategies (no suppression, partial suppression and full suppression) influences landscape patterns.

Study Areas

Each of our three study areas focuses on a large fire event in mixed conifer forests of western North America. These fires were part of regional fire years and were noteworthy for the size and the severity of the event. Study areas are within mountainous, forested landscapes with mixed conifer forests of lodgepole pine (Pinus contorta), subalpine fir (Abies lasiocarpa), and Engelmann spruce (Picea engelmannii). In each study area, a recent, large wildfire event interfaced with past wildfires.

Tripod

The 2006 Tripod Complex burned over 70,000 ha of the Okanogan-Wenatchee National Forest (MTBS 2010; Prichard and Kennedy 2014). Approximately 65% of the area burned as a stand replacement event. The study area supports a mix of vegetation types from low-elevation ponderosa pine (Pinus ponderosa) and Douglas-fir (Pseudotsuga menziesii) forests to high-elevation mixed conifer codominated by Engelmann spruce, subalpine fir, and lodgepole pine. At the highest elevations within this study area, forests yield to subalpine parklands of whitebark pine (Pinus albicaulis) and subalpine larch (Larix lyallii).

East Zone

In 2007, the East Zone Complex fires burned over 128,000 ha on the Boise and Payette National Forests in central Idaho (MTBS 2010; Hudak et al. 2011) and was active concurrently with adjacent large fires including the 128,000-ha Cascade Complex to the south and 40,000-ha Rattlesnake Complex to the North. The East Zone Complex study area was selected because it was in the center of the two other fires and supports a wide range of forest types and elevations from subalpine forests and meadows at high elevation to lower tree line dominated by ponderosa pine woodlands. Four State and Transition Models (STMs) were developed to represent low elevation dry and moist mixed conifer forests and high elevation cold dry and cold-moist forests.

Kootenay

The 17,000-ha 2003 Kootenay Fire Complex was one of the largest fire events to have occurred in the Canadian Rockies in the past century and burned within Kootenay National Park. The study area, located in the Canadian Rockies within southeastern British Columbia, is dominated by high elevation mixed conifer forests of Engelmann spruce, subalpine fir and lodgepole pine. Over 75% of the area burned at moderate to high severity. Pre-fire fuel complexes were comprised of mature mixed-conifer forests of lodgepole pine, Engelmann spruce, and subalpine fir. A striking feature of the post-burn landscape is the nearly uniform tree stand replacement within the burned area. Two STMs were developed for the Kootenay study area, representing cold and dry landscapes on exposed ridges and steep slopes and cold and moist landscapes on sites with greater moisture and higher productivity.

Research Questions

Question 1

Did past wildfires influence the spread and the severity of subsequent large fires?

Findings:

1. Past wildfires areas had significantly lower burn severity than surrounding areas.
   • In Tripod and Kootenay landscapes, past burn severity was a strong predictor of subsequent burn severity with the highest reductions where pixels previously burned in high severity fire events.
   • In Cascade, pixels that had previously burned in low severity fire events actually had the highest reduction in reburn severity, likely reflecting that flammable shrub fields regenerated in this study area following high-severity fire and then reburned at high severity.
2. Across all study areas, final models of burn severity included past fire effects (severity and distance from edge), weather, vegetation and landform.
   • Even though all 3 study areas burned within large fire events associated with regional drought years and extreme fire weather conditions, we found a mix of top-down (weather) controls and bottom-up (fuels, vegetation and landform) controls on burn severity.
   • Fires generally burned more severely on extreme weather days associated with high temperatures and low relative humidity.
   • Vegetation was a strong predictor of burn severity with higher burn severity in dense, closed canopy mixed conifer forests than low-elevation, more open ponderosa pine and moist riparian forest types.
   • Burn severity was also strongly related to landform with a trend toward higher severity on steeper slopes and increased severity with elevation until fires reached treeline.
3. Recent fires were generally barriers to fire spread within 5-7 years following the previous fire.
   • Each fire-on-fire interaction has a unique landscape and fire weather context.
   • Some recent small fires fully reburned in the Tripod landscape, but large fires were complete barriers.
   • Most recent fires in the East Zone and Cascades fires were complete barriers. However, a few fires reburned and appeared to have burned on slopes directly exposed to a head fire.
   • Post-fire landscapes in the Kootenay study area of interior British Columbia often develop continuous grass fuels that support reburns between 2-5 years following stand-replacing fire events.
     

     Percent area reburned in the Cascade, East Zone and Tripod Complex fires. Although some previous fires reburned within 5 years, most had low amounts of reburn. Area reburned generally increases over time.

Question 2

How do past wildfires influence or inform management strategies for subsequent wildfires?

The primary objective of this study was to evaluate how fire exclusion has contributed to altered vegetation and fuel patterns and to evaluate hypothetical landscapes under comparative fire management strategies.

Reburn Simulation Tool

The Reburn Simulation tool is an iterative, geospatial modeling tool that uses historical ignition and weather data to evaluate potential burn mosaics compared to actual pre-wildfire landscapes under different wildfire management strategies. The tool iteratively models fire spread using historical fire start and weather data for each recorded fire start. The Tripod landscape, for example, begins with recorded fire starts in 1940. The simulation tool starts with a base landscape and models fire spread for the suppressed fire starts in 1940 and then uses a set of State and Transition Models created for each study area to repopulate landscapes based on predicted flame lengths (converted to low, moderate and high severity effects) and vegetation and fuel succession over time.

State and Transition Models

As part of a study on burn mosaics and their effect on subsequent wildfires, we developed a series of State and Transition Models (STMs) to represent vegetation, fuels and fire dynamics in mixed conifer forests of western North America (pdf – STM guide). Although the three study areas are geographically distinct, they share similar vegetation types and historically complex mixed-severity fire regimes. Because fire is a dominant driver of vegetation within each of the three study areas, the fire and vegetation pathways reflect how fire events can reset succession back to stand initiation from high severity, stand replacement events or cause successional trajectories to branch under low to moderate severity events. The STMs were developed to support simulation modeling of vegetation and fire dynamics in western US and Canada, but the STMs have many potential applications, including translation to early and late-successional wildlife habitat, carbon stores and wildland fire emissions evaluations.

Sample cold dry STM for the East Zone study area, representing vegetation and fuel dynamics in Engelmann spruce, subalpine fir and lodgepole pine forests under low, moderate and high-severity fire pathways.

Sample photos of the Cold Dry Conifer fire exclusion pathway (A)

Sample photos of the Dry Mixed Conifer fire exclusion pathway (A)

Comparative Landscapes under Alternative Fire Management Strategies

Using simulation modeling with historical ignition and weather data, we evaluated potential burn mosaics compared to actual pre-wildfire landscapes under four management strategies:

• Complete absence of fire -- no fires
• Modern Suppression -- only fires that escape suppression
• Partial Suppression -- managed wildfires in the late-summer and fall fire seasons and escaped wildfires
• No Suppression – all ignitions that meet thresholds to burning

The simulation modeling results from our study offer a unique perspective of the long-term consequences of our wildfire management decisions – in particular, the implications of fire management decisions for future wildfire events. Of our four scenarios, the No Fire and Modern Suppression scenarios represent “boom and bust” landscapes in which continuous mature forests are capable of supporting large fire spread. The partial wildfire and no suppression landscapes have finer-grained patch mosaics and would have presented a markedly different landscape to fire managers in the 2006 Tripod Fire. Specifically, the partial and no suppression landscapes support a much more diverse landscape that is less susceptible to large, stand-replacing fire events and supports a wide range of forest ages.

• Complete absence of fire (no fire). The resulting landscape of the absence of fires from 1940 to 2006 reflects the maturation of forests and relatively homogenous landscapes of young- and old-forest multistory structures. This landscape bears a marked resemblance to the actual pre-fire Tripod landscape and the highly contagious patterns of vegetation and fuels that supported the actual wildfire event.
• Modern Suppression. The modern fire suppression scenario was designed to represent contemporary wildland fire management in which only the fires that escape suppression (the 2-3% of fires of that burn under extreme fire weather and cannot be suppressed) were allowed to burn. In general, results demonstrate a general infilling of the landscape with more mature forests prior to 2006 similar to the no-fire scenario.
• Partial Suppression. The partial suppression scenario allowed for managed wildfires in the late-summer and fall fire seasons and escaped wildfires. Simulations of the partial suppression scenario generally demonstrate finer-grain landscape mosaics at lower elevations that support dry, mixed conifer forests (southern portion of study area) and a mix of small and large patches at higher elevations (northern portion of study area).
• No suppression. In this frequent-fire scenario, the landscape supports a relatively low percentage of mature forest and had the finest patchwork of vegetation of any of the scenarios. Patches of young forest multistory and old forest multistory were generally surrounded by recent burns (black pixels) and regenerating forest.

Management Applications Implications for Lynx Habitat

Canada lynx were federally listed as a threatened species in 2000. Historically, north-central Washington State contained some of the largest contiguous blocks of habitat for Canada lynx in the United States. However, loss and fragmentation of habitat from wildfires has greatly reduced viable habitat for lynx.

We are using the Reburn project to link our burn mosaic modeling under contrasting wildfire management scenarios to evaluate potential consequences for lynx habitat.

Lynx surveys have revealed limited use of areas that recently burned as moderate and high severity. Regenerating forests 2-4 decades following fire provide high quality hare/lynx habitat. Old forests with canopy gaps provide understory cover and downed logs used as habitat for hare/lynx, including denning (not limited to old forests).

Historically, fire of various ages, provided for a mix of forest age classes and habitat components for lynx/hares across sub-boreal forest landscapes. Using results from the Reburn project, we are evaluating how suitable lynx habitat is maintained over time in the 4 scenarios. Although the no-fire and modern suppression scenarios provide plenty of late-successional habitat, large fires can eliminate much of this habitat. Alternative scenarios such as the partial wildfire scenario maintain lower amounts of suitable habitat for lynx but would likely be more resilient to future wildfires and may represent a more realistic carrying capacity for this threatened species.

Publications

2023

Povak, N.A., Hessburg, P.F., Salter, R.B., Gray, R.W., and Prichard, S.J. 2023. System-level feedbacks of active fire regimes in large landscapes. Fire Ecology 19:45. https://doi.org/10.1186/s42408-023-00197-0

Prichard, S.J., Salter, R.B., Hessburg, P.F., Povak, N., and Gray, R.W. 2023. The REBURN model: simulating system-level forest succession and wildfire dynamics. Fire Ecology 19:38. https://doi.org/10.1186/s42408-023-00190-7

In prep

Gaines, W.L., Hessburg, P.F., Lyons, A.L., Salter, R.B., Prichard, S.J., VanBianchi, C. and Hodges, K. In prep. Synergistic effects of climate change, large wildfires, and past management challenge the survival of an iconic cat: Canada lynx in the North Cascades, USA. Frontiers in Ecology and Environment.

Povak, N., Salter, B., Hessburg, P., Prichard, S.J. and Gray, R.W. In prep. Fire, fuel and vegetation dynamics – modeling wildfire and fuel succession in fire-prone landscapes. Part II: simulation modeling.

Povak, N., Salter, R.B., Prichard, S.J. and Gray, R.W.  In prep. Landscape mosaics under comparative wildfire management strategies in semi-arid forest landscapes of western North America.

Prichard, S.J., et al. In prep. Wildfires as fuel treatments – burn mosaics and wildfire management. Fire Management Today (draft version). View PDF (1.1 MB)

Prichard, S.J., Gray, B., Hessburg, P., Povak, N. and Salter, B. In prep. State and transition models of semi-arid forest landscapes in western North America: fire and fuel pathways. (Online handbook) View PDF (2.9 MB)

Prichard, S.J., Gray, R.W., Salter, B., Hessburg, P. and Povak, N. In prep. Fire, fuel and vegetation dynamics – modeling wildfire and fuel succession in fire-prone landscapes. Part I: state and transition models.

Older

Prichard, S.J., Stevens-Rumann, C.S., Hessburg, P. 2017. Tamm Review: Shifting global fire regimes: lessons from reburns and research needs. Forest Ecology and Management 396: 217-233. https://www.fs.fed.us/pnw/pubs/journals/pnw_2017_prichard001.pdf

Stevens-Rumann, C.S., Prichard, S.J., and Morgan, P. 2014. The Effect of Previous Wildfires on Subsequent Wildfire Behavior and Post Wildfire Recovery. Northern Rockies Fire Science Network Science Review No. 1. https://www.nrfirescience.org/sites/default/files/NRFSNSciReview1_RepeatFires.pdf

Stevens-Rumann, C.S., Prichard, S.J., Strand, E.K., Morgan, P. 2016. Prior wildfires influence burn severity of subsequent large fires. Canadian Journal of Forest Research 46 (11): 1375-1385. DOI: 10.1139/cjfr-2016-0185. https://www.researchgate.net/publication/308795672_Prior_wildfires_influence_burn_severity_of_subsequent_large_fires

Conference presentations

Laflamme, J., Hessburg, P., Salter, B., Prichard, S., and Gray, B. 2023. Understanding landscape-level feedbacks between fire and forest management in British Columbia’s Okanagan Region under historical and future climate scenarios. International Association of Landscape Ecology annual meeting, April 2023, Online. View PDF (1.8 MB)

Prichard, S.J. Fire on our Side Special Session, November 19, 2015. Past burn mosaics in the North Cascades Mountains: implications for wildland fire management. 6th International Conference on Fire Ecology and Management, San Antonio, TX.

Prichard, S.J. and Stevens-Rumann, C. Special Session, November 16, 2015. Wildland fire-on-fire interactions: A review of fire-prone ecosystems and implications under a changing climate. 6th International Conference on Fire Ecology and Management, San Antonio, TX.

Stevens-Rumann CS, Morgan P. 2017. Mosaic landscapes: what mixed severity fires, repeated fires and the subsequent landscapes tell us about habitat. The Wildlife Society, September 2017, Albuquerque, NM.

Stevens-Rumann CS, Prichard S, Morgan P, Higuera P, Harvey B. 2016. Wildfires for better or worse? Natural Areas Association Conference, October 2016, Davis, CA.

Stevens-Rumann CS, Prichard S, Morgan P. 2015. The evaluation of burn mosaics on subsequent wildfire burn severity and postfire effects. AFE Fire Congress, November 2015, San Antonio, TX.