Research and Publications
View article.
We generated burn severity spectral index values for a dataset of Composite Burn Index (CBI) field plots across the conterminous US. The effectiveness of offset corrections was tested across image selection techniques, spectral indices, offset generation methods and burn perimeter sources. We assessed the influence of offset corrections on the modeled relationship with CBI, agreement between burn severity thresholds and potential bias. Applying offset corrections consistently improved the modeled relationship with CBI by addressing extreme outlier severity values. However, automated offset corrections had the potential to introduce bias, systematically lowering severity values and reducing correspondence with observed burn severity categories. Offset corrections offer benefits but also present trade-offs to accurately representing remotely sensed burn severity.
View article.
To assess the role of current and source environments in explaining variation in flowering phenology of Bromus tectorum, an invasive annual grass, we conducted a replicated common garden experiment using 92 genotypes collected across western North America. Replicates of each genotype were planted in two densities (low = 100 seeds/1 m², high = 100 seeds/0.04 m²) under two different temperature treatments (low = white gravel; high = black gravel; 2.1°C average difference) in a factorial design, replicated across four common garden locations in Idaho and Wyoming, USA. We tested for the effect of current environment (i.e., density treatment, temperature treatment, and common garden location), source environment (i.e., genotype source climate), and their interaction on each plant’s flowering phenology. Flowering timing was strongly influenced by a plant’s current environment, with plants that experienced warmer current climates and higher densities flowering earlier than those that experienced cooler current climates and lower densities. Genotypes from hot and dry source climates flowered consistently earlier than those from cool and wet source climates, even after accounting for genotype relatedness, suggesting that this genetically based climate cline is a product of natural selection. We found minimal evidence of interactions between current and source environments or genotype-by-environment interactions. Phenology was more sensitive to variation in the current climate than to variation in source climate. These results indicate that cheatgrass phenology reflects high levels of plasticity as well as rapid local adaptation. Both processes likely contribute to its current success as a biological invader and its capacity to respond to future environmental change.
View article.
Local adaptation may facilitate range expansion during invasions, but the mechanisms underlying successful invasions remain unclear. Cheatgrass (Bromus tectorum), native to Eurasia and Africa, has invaded globally, with severe impacts in western North America. We aim to identify mechanisms and consequences of local adaptation in the North American cheatgrass invasion. We sequence 307 range-wide genotypes and conduct controlled experiments. We find that diverse lineages invaded North America, where long-distance gene flow is common. Nearly half of North American cheatgrass comprises a mosaic of ~19 locally adapted, near-clonal genotypes, each seemingly very successful in a different part of North America. Additionally, ancestry, phenotype, and allele frequency-environment clines in the native range predict those in the invaded range, indicating pre-adapted genotypes colonized different regions. Common gardens show directional selection on flowering time that reverse between warm and cold sites, potentially maintaining clines. In the USA Great Basin, genomic predictions of strong local adaptation identify sites where cheatgrass is most dominant. Our results indicate that multiple introductions and migration within the invaded range fuel local adaptation and success of cheatgrass in western North America. Understanding how environment and gene flow shape adaptation and invasion is critical for managing ongoing invasions.
View article.
Forest growth, fire behavior, fire spread and emissions models were used to simulate fuel treatments and their potential impacts. The ‘underburn only’ and ‘thin from below + pile burn’ treatments had a minimum annual fire probability (AFP) 5–35% lower than other treatment types to achieve reduced GHG emissions. When AFP was high, the ‘stand density index (SDI) thin + underburn’ treatment reduced GHG emissions 13–54% more than the next best treatment. AFP, forest type and initial hazard level should be primary considerations when selecting a fuel treatment type for reducing future GHG emissions.
View article.
We used 26,907 California’s wildfire ignitions to run automatic simulations. The Fire Behavior Index (FBI), Terrain Difficulty Index (TDI) and IAA were determined using size-based and fire behavior outputs from each simulation. Initial attack success was evaluated by comparing simulations with real fire sizes. Binary models were calibrated and validated to predict success based on IAA, FBI and TDI, and suppression response time.
The IAA effectively identified fires exceeding suppression capacity. Higher levels of IAA, FBI and TDI were associated with reduced success odds, IAA=5 giving a 90% decrease in the odds of initial attack success. Response time and its interaction with terrain difficulty were also influential. The IAA is a powerful index to feed DSSs, prioritizing fire response and predicting the probability of control at a small size.
View article.
We measured forb composition and other community characteristics at different scales in 17 sites across eastern Oregon with paired plots across seeding boundaries. We quantified differences associated with seeding overall. We also investigated the potential for seeding outcomes to relate to biotic interactions based on abundance of major functional groups, species traits related to competition, microsite features relating to competition, and soil texture. Finally, we tested the generality of our paired plot results across the study area by accounting for a broader range of management and environmental factors with a matching analysis using a large independent vegetation dataset. Paired plot comparisons suggested negative relationships between seeding and forb density. Biotic interactions with major functional groups were not strongly associated with forb community differences due to seeding, with the exception of a negative relationship between native annual forb abundance and increasing invasive annual grass cover. Few biotic interactions were related to seeding outcomes, though native annual forb abundance was negatively associated with higher invasive annual grass cover. Traits related to competition were not strongly related to species seeding response. Soil texture affected several forb community characteristics, including sometimes altering seeding outcomes. Increasing sand content was associated with lower perennial forb diversity and abundance, and annual forb richness decreased with higher sand content in seeded, but not unseeded, sites. The regional-level matching analysis, which controlled for the effects of management and environmental factors, generally supported the paired plot comparisons by showing that crested wheatgrass seeding was associated with lower invasive annual grass cover and lower total and perennial forb cover and richness.
View article.
We measured herbaceous biomass response to cattle grazing spanning 18 yr (2007–2024) on burned Wyoming big sagebrush steppe in southeastern Oregon. Treatments were applied in a randomized complete block design, including no grazing on burned (nonuse) and unburned (control) sagebrush steppe; and cattle grazing at low (low), moderate (moderate), and high (high) stocking. All grazed treatments were by deferred rotation. Deferred rotation consisted of grazing during the active growing season (mid-May–early June) once every 3 yr followed by 2 yr of grazing during summer herbaceous dormancy (July, August, or September). Herbage was sorted by herbaceous functional group, which included an early season bunchgrass, tall perennial bunchgrasses, perennial forbs, cheatgrass, and annual forbs. Both standing crop and annual net primary production (ANPP, current year’s growth) of functional groups were evaluated by repeated measures analysis. Standing crop decreased as grazing intensity increased but recovered with 1 or 2 yr of grazing rest. Herbaceous functional group ANPP did not differ among the burned treatments (grazed and nonuse), and total and perennial bunchgrass production were all greater than the control. Grazing intensity in the deferred rotation program did not affect long-term ANPP. Annual weather events account for ANPP variability measured for the various grazed and ungrazed treatments.
View article.
We compared vegetation recovery spanning 20 yr following prescribed fire on mid-succession and late-succession western juniper woodlands on Steens Mountain, Oregon. Our objective was to evaluate vegetation dynamics between early (first decade) and later successional (second decade) time periods after fire. The first decade after fire vegetation on burned mid-succession sites were codominated by native herbaceous perennials and sprouting shrub species and on late-succession sites vegetation was codominated by nonnative cheatgrass and snowbrush. During the second decade after fire, vegetation composition converged and both mid-succession and late-succession sites were codominated by herbaceous perennials, mountain big sagebrush, round-leaf snowberry, and snowbrush. Herbaceous and shrub vegetation composition of both burned woodland phases proved to be highly resilient to fire, the difference was that native shrub-herbaceous recovery on late-succession sites required about twice as much time as mid-succession sites. The resilience of both mid-succession and late-succession woodland sites was likely a product of ecological site characteristics (e.g., elevation and precipitation zone) that affords a competitive advantage for native perennial species over invasive annuals.
View article.
The Bureau of Land Management (BLM) plans to expand its network of LFBs in the Great Basin by over 17 000 km. However, uncertainties remain regarding their effectiveness in reducing wildfire-related impacts. To address this knowledge gap, we estimate avoided wildfire costs attributable to fuel breaks in the Twin Falls BLM District of south-central Idaho. Our analysis focuses on the 2019 Pothole fire, which was contained in part due to the presence of LFBs. By developing a counterfactual simulated scenario in which the fire did not intersect the fuel breaks and using historic data on suppression expenditures, postfire rehabilitation costs, and grazing-related forage losses, we estimate the net economic benefits associated with fuel break presence. This case study provides actionable insights for land managers by quantifying the potential cost savings from fuel break infrastructure. Our findings indicate that in the northern Great Basin, LFBs may significantly reduce wildfire management costs, supporting their strategic deployment as part of a broader landscape-scale fire mitigation approach.
View article.
This study investigates the key drivers influencing prescribed fire effects across 16 sites in northern and central California, with particular emphasis on how operational decisions by fire practitioners shape burn outcomes. Data from the California Prescribed Fire Monitoring Program revealed that prescribed fires reduced total fuel loads by an average of 60 %, with greater consumption of postfrontal smoldering fuels (coarse fuels, 65 %) compared to frontline spreading fuels (fine fuels, 49.0 %).
Crown scorch height showed a strong relationship to crown base height (R2 = 0.37–0.86), suggesting that practitioners use crown base height as a visual indicator to control fireline intensity and avoid crown damage. This relationship may partially explain fuel consumption patterns, as crown avoidance strategies can influence fire behavior and intensity. Additionally, we documented a compensatory relationship between live and dead fuel moisture content across burn seasons, indicating that practitioners strategically select burning windows that maintain fireline intensity within controllable parameters regardless of season.
Our findings demonstrate that human decisions fundamentally modify prescribed fire behavior to maintain safety parameters, often constraining outcomes to conservative ranges that may compromise treatment effectiveness. Understanding and accounting for these human factors is crucial to encouraging a more effective use of prescribed fires in the future. We recommend that future research explicitly include operational parameters and practitioner decision-making processes in assessing prescribed fire science, balancing safety considerations with goals for ecological restoration.