Tania Schoennagel

The Interaction of Fire, Fuels, and Climate across Rocky Mountain Forests

Abstract Understanding the relative influence of fuels and climate on wildfires across the Rocky Mountains is necessary to predict how fires may respond to a changing climate and to define effective fuel management approaches to controlling wildfire in this increasingly populated region. The idea that decades of fire suppression have promoted unnatural fuel accumulation and subsequent unprecedentedly large, severe wildfires across western forests has been developed primarily from studies of dry ponderosa pine forests. However, this model is being applied uncritically across Rocky Mountain forests (e.g., in the Healthy Forests Restoration Act). We synthesize current research and summarize lessons learned from recent large wildfires (the Yellowstone, Rodeo-Chediski, and Hayman fires), which represent case studies of the potential effectiveness of fuel reduction across a range of major forest types. A “one size fits all” approach to reducing wildfire hazards in the Rocky Mountain region is unlikely to be effective and may produce collateral damage in some places.

Learning to coexist with wildfire

The impacts of escalating wildfire in many regions — the lives and homes lost, the expense of suppression and the damage to ecosystem services — necessitate a more sustainable coexistence with wildfire. Climate change and continued development on fire-prone landscapes will only compound current problems. Emerging strategies for managing ecosystems and mitigating risks to human communities provide some hope, although greater recognition of their inherent variation and links is crucial. Without a more integrated framework, fire will never operate as a natural ecosystem process, and the impact on society will continue to grow. A more coordinated approach to risk management and land-use planning in these coupled systems is needed.

Managing fire-prone forests in the western United States

The management of fire-prone forests is one of the most controversial natural resource issues in the US today, particularly in the west of the country. Although vegetation and wildlife in these forests are adapted to fire, the historical range of fire frequency and severity was huge. When fire regimes are altered by human activity, major effects on biodiversity and ecosystem function are unavoidable. We review the ecological science relevant to developing and implementing fire and fuel management policies for forests before, during, and after wildfires. Fire exclusion led to major deviations from historical variability in many dry, low-elevation forests, but not in other forests, such as those characterized by high severity fires recurring at intervals longer than the period of active fire exclusion. Restoration and management of fire-prone forests should be precautionary, allow or mimic natural fire regimes as much as possible, and generally avoid intensive practices such as post-fire logging and planting.

ENSO AND PDO VARIABILITY AFFECT DROUGHT-INDUCED FIRE OCCURRENCE IN ROCKY MOUNTAIN SUBALPINE FORESTS

Understanding the effect of variation in climate on large-fire occurrence across broad geographic areas is central to effective fire hazard assessment. The El Nino- Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO) affect winter tem- perature and precipitation regimes in western North America through mid-latitude tele- connections. This study examines relationships of ENSO and the PDO to drought-induced fire occurrence in subalpine forests of three study areas across the Rocky Mountains: Jasper National Park (JNP, northern Rockies), Yellowstone National Park (YNP, central Rockies) and Rocky Mountain National Park (RMNP, southern Rockies) over the 1700-1975 period. Large-scale climatic anomalies captured by ENSO (NINO3) and PDO indices had differ- ential effects on large-fire occurrence across the study areas. Superposed epoch analysis (SEA) showed that large fires in RMNP occurred during extreme La Nina years, while the PDO, although predominantly negative during fire years, did not depart significantly from the mean. In YNP and JNP, neither ENSO nor PDO indices were significantly different from the mean during large-fire years, although fires tended to occur during El Nino and positive PDO years. Constructive phases (years of combined warm (positive) or cool (neg- ative) phases) of ENSO and the PDO were significantly associated with large-fire occurrence across the Rockies, even though these large-scale climatic anomalies were not significant when considered singly in SEAs. Combined warm phases (positive PDO during El Nino) co-occurred with large fires in the central and northern Rockies, while the combined cool phases (negative PDO during La Nina) appeared to promote large fires in the southern Rockies. Almost 70% of large fires in RMNP burned during La Nina events that coincided with a negative PDO, although these phases co-occurred during only 29% of the 1700- 1975 period. Spatial teleconnection patterns between drought, PDO and ENSO across west- ern North America independently support the sign and strength of relationships between these climatic anomalies and subalpine fire occurrence along a broad north-south gradient of the Rockies. Forecasts of ENSO that are dependent on the expected PDO phase suggest promise for fire hazard prediction across the West.

Spatiotemporal patterns of mountain pine beetle activity in the southern Rocky Mountains.

The current mountain pine beetle (MPB; Dendroctonus ponderosae) outbreak in the southern Rocky Mountains has impacted approximately 750 000 ha of forest. Weather and habitat heterogeneity influence forest insect population dynamics at multiple spatial and temporal scales. Comparison of forest insect population dynamics in two principal host species may elucidate the relative contribution of weather and landscape factors in initiating and driving extensive outbreaks. To investigate potential drivers of the current MPB outbreak, we compared broadscale spatiotemporal patterns of MPB activity in lodgepole pine (Pinus contorta) and ponderosa pine (Pinus ponderosa) from 1996 to 2010 in Colorado and southern Wyoming with regional weather fluctuations, and then tracked the annual meso-scale progression of the epidemic in lodgepole pine with respect to weather, topographic, previous MPB activity, and forest stand attributes. MPB activity in lodgepole pine compared to ponderosa pine showed higher magnitude and extent of spatial synchrony. Warm temperatures and low annual precipitation favorable to beetle populations showed high regional synchrony across areas of both pine species, suggesting that habitat interacts with weather in synchronizing MPB populations. Cluster analysis of time series patterns identified multiple, disjunct locations of incipient MPB activity (epicenters) in lodgepole pine, which overlapped an earlier 1980s MPB outbreak, and suggests a regional trigger (drought) across this homogenous forest type. Negative departures from mean annual precipitation played a key role in subsequent spread of MPB outbreak. Development of the outbreak was also associated with lower elevations, greater dominance by lodgepole pine, stands of larger tree size, and stands with higher percentage canopy cover. After epidemic levels of MPB activity were attained, MPB activity was less strongly associated with stand and weather variables. These results emphasize the importance of considering differences in patterns of MPB dynamics for different host pine species even under similar regional-scale weather variation and the nonstationarity of outbreak dynamics over time.

THE INFLUENCE OF FIRE INTERVAL AND SEROTINY ON POSTFIRE LODGEPOLE PINE DENSITY IN YELLOWSTONE NATIONAL PARK

The time interval between stand-replacing fires can influence patterns of initial postfire succession if the abundance of postfire propagules varies with prefire stand age. We examined the effect of fire interval on initial postfire lodgepole pine ( Pinus contorta var. latifolia Engelm.) density in Yellowstone National Park (YNP) following the 1988 fires. We asked whether postfire propagule abundance, measured as prefire percent serotiny, varied with fire interval and could explain patterns in postfire succession. The response of lodgepole pine density to variation in fire interval was explained by spatial and temporal variation in prefire serotiny. At low elevations, postfire lodgepole pine recruitment correlated strongly with prefire percent serotiny, which varied nonlinearly with prefire stand age. As a result, postfire lodgepole pine densities varied nonlinearly with fire interval. In contrast, at high elevations serotiny was low, varied little with stand age and did not influence postfire lodgepole pine densities, although, fire interval was still a significant predictor of postfire densities. At high elevations, fire interval varied nonlinearly with postfire lodgepole densities, presumably due to the temporal variation in propagule abundance from open cones in adjacent unburned stands. Temporal variation in stand-level serotiny at low ele- vations was best explained by age of individual trees. Logistic regression indicated that trees expected to be serotinous had a low probability of exhibiting serotiny at a young age, with increasing probability as trees matured up to 140 yr. This increase in serotiny with tree age likely accounts for the initial increase in stand-level percent serotiny with stand age at low elevations. The spatial variation in serotiny was correlated with variation in historical fire regimes. Fire interval models derived from lower elevations in YNP indicate that fire occurred historically at 135-185-yr intervals, whereas at higher elevations fires occurred at 280-310-yr intervals. The spatial patterns of serotiny appear to have been influenced by variability in historical fire regimes across the Yellowstone landscape, which has conditioned contemporary successional responses to disturbance.

Implementation of National Fire Plan treatments near the wildland–urban interface in the western United States

Because of increasing concern about the effects of catastrophic wildland fires throughout the western United States, federal land managers have been engaged in efforts to restore historical fire behavior and mitigate wildfire risk. During the last 5 years (2004–2008), 44,000 fuels treatments were implemented across the western United States under the National Fire Plan (NFP). We assessed the extent to which these treatments were conducted in and near the wildland–urban interface (WUI), where they would have the greatest potential to reduce fire risk in neighboring homes and communities. Although federal policies stipulate that significant resources should be invested in the WUI, we found that only 3% of the area treated was within the WUI, and another 8% was in an additional 2.5-km buffer around the WUI, totaling 11%. Only 17% of this buffered WUI is under federal ownership, which significantly limits the ability of federal agencies to implement fire-risk reduction treatments near communities. Although treatments far from the WUI may have some fire mitigation benefits, our findings suggest that greater priority must be given to locating treatments in and near the WUI, rather than in more remote settings, to satisfy NFP goals of reducing fire risk to communities. However, this may require shifting management and policy emphasis from public to private lands.

Spatial variability in wildfire probability across the western United States

Despitegrowingknowledgeoffire-environmentlinkagesinthewesternUSA,obtainingreliableestimatesof relativewildfirelikelihoodremainsaworkinprogress.Thepurposeofthisstudyistouseupdatedfireobservationsduring a 25-year period and a wide array of environmental variables in a statistical framework to produce high-resolution estimatesofwildfireprobability.UsingtheMaxEntmodellingtechnique,point-sourcefireobservationsthatweresampled from area burned during the 1984-2008 time period were related to explanatory variables representing ignitions, flammable vegetation (i.e. fuels), climate and topography. Model results were used to produce spatially explicit predictions of wildfire probability. To assess the effect of humans on the spatial patterns of wildfire likelihood, we built an alternative model that excluded all variables having a strong anthropogenic imprint. Results showed that wildfire probabilityinthewesternUSAisfarfromuniform,withdifferentareasrespondingtodifferentenvironmentaldrivers.The effect of anthropogenic factors on wildfire probability varied by region but, on the whole, humans appear to inhibit fire activity in the western USA. Our results not only provide what appear to be robust predictions of wildfire likelihood, but also enhance understanding of long-term controls on wildfire activity. In addition, our wildfire probability maps provide betterinformationforstrategicplanningofland-managementactivities,especiallywherefireregimeknowledgeissparse. Additional keywords: climate, fuels, ignitions, MaxEnt algorithm, spatial modelling, topography.

Landscape heterogeneity following large fires: insights from Yellowstone National Park, USA

We characterised the remarkable heterogeneity following the large, severe fires of 1988 in Yellowstone National Park (YNP), in the northern Rocky Mountains, Wyoming, USA, by focussing on spatial variation in post-fire structure, composition and ecosystem function at broad, meso, and fine scales. Ecological heterogeneity at multiple scales may enhance resilience to large, severe disturbances by providing structural, biological and functional redundancy. Post-fire heterogeneity in stand age, coarse wood abundance, microbial and understorey communities reflected interactions between existing pre-fire patterns and fire severity at different scales, suggesting that environmental context plays an important role in successional responses to large fires. In response to these post-fire patterns, heterogeneity in carbon (C) and nitrogen (N) storage, N mineralisation, decomposition, and productivity was also evident at multiple scales and may confer resiliency to large fires. For example, at broad scales, C storage in YNP appears resistant to changes in age-class structure associated with large stand-replacing fires. In summary, the YNP landscape is recovering rapidly from the 1988 fires through natural mechanisms, owing to the abundance and spatial heterogeneity of post-fire residuals, but other systems with fewer biotic legacies may be less resilient to such large, severe fires.

Effects of Mountain Pine Beetle on Fuels and Expected Fire Behavior in Lodgepole Pine Forests, Colorado, USA

In Colorado and southern Wyoming, mountain pine beetle (MPB) has affected over 1.6 million ha of predominantly lodgepole pine forests, raising concerns about effects of MPB-caused mortality on subsequent wildfire risk and behavior. Using empirical data we modeled potential fire behavior across a gradient of wind speeds and moisture scenarios in Green stands compared three stages since MPB attack (Red [1–3 yrs], Grey [4–10 yrs], and Old-MPB [∼30 yrs]). MPB killed 50% of the trees and 70% of the basal area in Red and Grey stages. Across moisture scenarios, canopy fuel moisture was one-third lower in Red and Grey stages compared to the Green stage, making active crown fire possible at lower wind speeds and less extreme moisture conditions. More-open canopies and high loads of large surface fuels due to treefall in Grey and Old-MPB stages significantly increased surface fireline intensities, facilitating active crown fire at lower wind speeds (>30–55 km/hr) across all moisture scenarios. Not accounting for low foliar moistures in Red and Grey stages, and large surface fuels in Grey and Old-MPB stages, underestimates the occurrence of active crown fire. Under extreme burning conditions, minimum wind speeds for active crown fire were 25–35 km/hr lower for Red, Grey and Old-MPB stands compared to Green. However, if transition to crown fire occurs (outside the stand, or within the stand via ladder fuels or wind gusts >65 km/hr), active crown fire would be sustained at similar wind speeds, suggesting observed fire behavior may not be qualitatively different among MPB stages under extreme burning conditions. Overall, the risk (probability) of active crown fire appears elevated in MPB-affected stands, but the predominant fire hazard (crown fire) is similar across MPB stages and is characteristic of lodgepole pine forests where extremely dry, gusty weather conditions are key factors in determining fire behavior.