Search

We report a fine-scale assessment of cross-boundary wildfire events for the western US. We used simulation modeling to quantify the extent of fire exchange among major federal, state, and private land tenures and mapped locations where fire ignitions can potentially affect populated places. We examined how parcel size affects wildfire transmission and partitioned the relative amounts of transmitted fire between human and natural ignitions. We estimated that 85 % of the total predicted wildfire activity, as measured by area burned, originates from four land tenures (Forest Service, Bureau of Land Management, private, and state lands) and 63 % of the total amount results from natural versus human ignitions. On average, one-third of the area burned by predicted wildfires was nonlocal, meaning that the source ignition was on a different land tenure. Land tenures with smaller parcels tended to receive more incoming fire on a proportional basis, while the largest fires were generated from ignitions in national parks, national forests, and public and tribal lands. Among the 11 western states, the amount and pattern of cross-boundary fire varied substantially in terms of which land tenures were mostly exposed, by whom, and to what extent. We also found spatial variability in terms of community exposure among states, and more than half of the predicted structure exposure was caused by ignitions on private lands or within the wildland–urban interface areas. This study addressed gaps in existing wildfire risk assessments that do not explicitly consider cross-boundary fire transmission and do not identify the source of fire. The results can be used by state, federal, and local fire planning organizations to help improve risk mitigation programs.

Climate change is altering wildfire and vegetation regimes in California’s forested ecosystems. Present day fires are seeing an increase in high burn severity area and high severity patch size. The ability to predict future burn severity patterns could better support policy and land management decisions. Here we demonstrate a methodology to first, statistically estimate individual burn severity classes at 30 meters and second, cluster and smooth high severity patches onto a known landscape. Our goal here was not to exactly replicate observed burn severity maps, but rather to utilize observed maps as one realization of a random process dependent on climate, topography, fire weather, and fuels, to inform creation of additional realizations through our simulation technique. We developed two sets of empirical models with two different vegetation datasets to test if coarse vegetation could accurately model for burn severity. While visual acuity can be used to assess the performance of our simulation process, we also employ the Ripley’s K function to compare spatial point processes at different scales to test if the simulation is capturing an appropriate amount of clustering. We utilize FRAGSTATS to obtain high severity patch metrics to test the contiguity of our high severity simulation. Ripley’s K function helped identify the number of clustering iterations and FRAGSTATS showed how different focal window sizes affected our ability to cluster high severity patches. Improving our ability to simulate burn severity may help advance our understanding of the potential influence of land and fuels management on ecosystem-level response variables that are important for decision-makers. Simulated burn severity maps could support managing habitat and estimating risks of habitat loss, protecting infrastructure and homes, improving future wildfire emissions projections, and better mapping and planning for fuels treatment scenarios.

Large, severe fires are becoming more frequent in many forest types across the western United States and have resulted in tree mortality across tens of thousands of hectares. Conifer regeneration in these areas is limited because seeds must travel long distances to reach the interior of large burned patches and establishment is jeopardized by increasingly hot and dry conditions. To better inform postfire management in low elevation forests of California, USA, we collected 5-yr postfire recovery data from 1,234 study plots in 19 wildfires that burned from 2004-2012 and 18 yrs of seed production data from 216 seed fall traps (1999-2017). We used these data in conjunction with spatially extensive climate, topography, forest composition, and burn severity surfaces to construct taxon-specific, spatially explicit models of conifer regeneration that incorporate climate conditions and seed availability during postfire recovery windows. We found that after accounting for other predictors both postfire and historical precipitation were strong predictors of regeneration, suggesting that both direct effects of postfire moisture conditions and biological inertia from historical climate may play a role in regeneration. Alternatively, postfire regeneration may simply be driven by postfire climate and apparent relationships with historical climate could be spurious. The estimated sensitivity of regeneration to postfire seed availability was strongest in firs and all conifers combined and weaker in pines. Seed production exhibited high temporal variability with seed production varying by over two orders of magnitude among years. Our models indicate that during droughts postfire conifer regeneration declines most substantially in low-to-moderate elevation forests. These findings enhance our mechanistic understanding of forecasted and historically documented shifts in the distribution of trees.

A novel approach is presented to analyze smoke exposure and provide a metric to quantify health-related impacts. Our results support the current understanding that managing low-intensity fire for ecological benefit reduces exposure when compared to a high-intensity full suppression fire in the Sierra Nevada of California. More frequent use of fire provides an opportunity to mitigate smoke exposure for both individual events and future emission scenarios. The differences in relative exposure between high-intensity, low-intensity, and prescribed burn were significant (P value

We developed and applied a wildfire simulation package in the Envision agent-based landscape modelling system. The wildfire package combines statistical modelling of fire occurrence with a high-resolution, mechanistic wildfire spread model that can capture fine scale effects of fire feedbacks and fuel management, and replicate restoration strategies at scales that are meaningful to forest managers. We applied the model to a landscape covering 1.2 million ha of fire prone area in central Oregon, USA where wildland fires are increasingly impacting conservation, amenity values and developed areas. We conducted simulations to examine the effect of human versus natural ignitions on future fire regimes under current restoration programs, and whether contemporary fire regimes observed in the past 20 years are likely to change as result of fire feedbacks and management activities. The ignition prediction model revealed non-linear effects of location and time of year, and distinct spatiotemporal patterns for human versus natural ignitions. Fire rotation interval among replicate simulations varied from 78 to 170 years and changed little over the 50-yr simulation, suggesting a stable but highly variable and uncertain future fire regime. Interestingly, the potential for fire-on-fire feedbacks was higher for human versus natural ignitions due to human ignition hotspots within the study area. We compare the methods and findings with other forest landscape simulation model (FLSM) studies and discuss future application of FLSMs to address emerging wildfire management and policy issues on fire frequent forests in the western US.

Shifting disturbance regimes can have cascading effects on many ecosystems processes. This is particularly true when the scale of the disturbance no longer matches the regeneration strategy of the dominant vegetation. In the yellow pine and mixed conifer forests of California, over a century of fire exclusion and the warming climate are increasing the incidence and extent of stand-replacing wildfire; such changes in severity patterns are altering regeneration dynamics by dramatically increasing the distance from live tree seed sources. This has raised concerns about limitations to natural reforestation and the potential for conversion to non-forested vegetation types, which in turn has implications for shifts in many ecological processes and ecosystem services. We used a California region-wide data set with 1,848 plots across 24 wildfires in yellow pine and mixed conifer forests to build a spatially explicit habitat suitability model for forecasting postfire forest regeneration. To model the effect of seed availability, the critical initial biological filter for regeneration, we used a novel approach to predicting spatial patterns of seed availability by estimating annual seed production from existing basal area and burn severity maps. The probability of observing any conifer seedling in a 60-m2 area (the field plot scale) was highly dependent on 30-yr average annual precipitation, burn severity, and seed availability. We then used this model to predict regeneration probabilities across the entire extent of a "new" fire (the 2014 King Fire), which highlights the spatial variability inherent in postfire regeneration patterns. Such forecasts of postfire regeneration patterns are of importance to land managers and conservationists interested in maintaining forest cover on the landscape. Our tool can also help anticipate shifts in ecosystem properties, supporting researchers interested in investigating questions surrounding alternative stable states, and the interaction of altered disturbance regimes and the changing climate.

Trail-based recreation is a popular use of public forests in the United States, and four types are common: all-terrain vehicle (ATV) riding, mountain biking, hiking, and horseback riding. Effects on wildlife, however, are controversial and often a topic of land use debates. Accordingly, we studied trail-based recreation effects on elk (Cervus canadensis), a wide-ranging North American ungulate highly sought for hunting and viewing on public forests, but that is sensitive to human activities, particularly to motorized traffic on forest roads. We hypothesized that elk would respond to trail-based recreation similarly to their avoidance of roads open to motorized traffic on public forests. We evaluated elk responses using a manipulative landscape experiment in a 1453-ha enclosure on public forest in northeast Oregon. A given type of recreation was randomly selected and implemented twice daily along 32 km of designated recreation trails over a five-day period, followed by a nine-day control period of no human activity. Paired treatment and control replicates were repeated three times per year for each recreation type during spring-fall, 2003–2004. During treatments, locations of elk and recreationists were simultaneously collected with telemetry units. Elk locations also were collected during control periods. Elk avoided the trails during recreation treatments, shifting distribution farther out of view and to areas farthest from trails. Elk shifted distribution back toward trails during control periods of no human activity. Elk avoided recreationists in real time, with mean minimum separation distances from humans that varied from 558 to 879 m among the four treatments, 2–4 times farther than elk distances from trails during recreation. Separation distances maintained by elk from recreationists also were 3–5 times farther than mean distances at which elk could be viewed from trails. Distances between elk and recreationists were highest during ATV riding, lowest and similar during hiking and horseback riding, and intermediate during mountain biking. Our results support the hypothesis that elk avoid trail-based recreation similarly to their avoidance of roads open to motorized traffic on public forests. Forest managers can use results to help optimize trade-offs between competing objectives for trail-based recreation and wildlife species like elk that are sensitive to human activities on public forests.

Recent mountain pine beetle outbreaks in whitebark pine forests have been extensive and severe. Understanding the climate influences on these outbreaks is essential for developing management plans that account for potential future mountain pine beetle outbreaks, among other threats, and informing listing decisions under the Endangered Species Act. Prior research has focused on one geographic region, but geographic variability in beetle and tree physiological responses to climate conditions have been documented. Here we evaluate geographic variability in climate influences on recent beetle outbreaks in whitebark pine and estimate future climate suitability for outbreaks across much of the range of whitebark pine in the western US. To accomplish these objectives, we developed and analyzed statistical models for three different geographic regions as well as a Westwide model, then applied the Westwide model to a suite of climate projections. The general patterns of climate-tree mortality relationships were similar across the three regions of our study. However, the relative importance of individual climate metrics preceding and during the recent outbreaks varied geographically because of the different climates in the regions. Winter minimum temperatures appeared to be limiting prior to outbreaks in the colder regions. All regions experienced low summer precipitation prior to or during outbreak initiation. Future climate suitability for beetle outbreaks is estimated to increase or remain stable in the coldest regions and decline slightly in the warmest region by the end of this century. Across the study area, projections of higher winter temperatures and decreased summer precipitation (with lower confidence than for temperatures) contribute to increased climate suitability for outbreaks, while projected higher fall/spring/summer temperatures contribute to decreased suitability. Some regional variability exists; in particular, the effect of winter warming is muted in the warmest region (Cascades) where winter temperatures appear to be less limiting. However, all regions are projected to experience fewer years with very low suitability, which commonly occurred prior to the recent outbreaks and may have limited beetle populations. Given the inherent uncertainty in climate projections and ecological responses to novel climates, management plans that incorporate sites that are expected to experience a range of expected future climate conditions might increase the chances of whitebark pine persistence in a warmer future.

The level of tree mortality and drought observed over the past decade in North America has been described as ‘unparalleled’ in our modern history, in particular in the Sierra Nevada, California. Forest managers could use early warning of where and how much tree mortality to expect in the very near future to plan and prioritize hazard tree removal, pest suppression activities, infer location of funding needs and fuels reduction treatments as well as access for firefighting. To answer these needs, we developed an empirically-based forecast model for expected tree mortality for an upcoming year based on (1) previous years’ tree mortality as observed in late summer; (2) previous years’ hydrologic year precipitation levels; and (3) site characteristics including amount of available host. Using this approach, initial forecasts for the next growing season can be developed by late fall for the following late summer. We demonstrated the application of this model by developing a forecast for the state of California for 2017. The explanatory variables in the California model accounted for ∼43% of the variability in tree mortality. Overall, the model missed forecasting high levels of mortality in approximately 5% of forested or woodland locations for the state of California. Locations with more mortality than expected in 2015 & 2016 were mostly associated with new outbreaks of insects; land use changes, and margins of prescribed- or wildfires not initially attributed. The forecasts may also be useful to natural resource and land managers in locating new outbreaks that may be attributed to novel behavior or exotic insects.

We report a fine scale assessment of cross-boundary wildfire events for the western US. We used simulation modeling to quantify the extent of fire exchange among major federal, state, and private land tenures and mapped locations where fire ignitions can potentially affect populated places. We examined how parcel size effects the wildfire transmission and partitioned the relative amounts of transmitted fire between human and natural ignitions. We estimated that almost 90 % of the total predicted wildfire activity as measured by area burned originates from four land tenures (Forest Service, Bureau of Land Management, private and State lands) and 63 % of the total amount results from natural versus human ignitions. On average, one third of the area burned by predicted wildfires was non-local, meaning that the source ignition was on a different land tenure. Land tenures with smaller parcels tended to receive more incoming fire on a proportional basis, while the largest fires were generated from ignitions in national parks, national forests, public and tribal lands. Among the 11 western States, the amount and pattern of cross-boundary fire varied substantially in terms of which land tenures were mostly exposed, by whom and to what fire sizes. We also found spatial variability in terms of community exposure among States, and more than half of the predicted structure exposure was caused by ignitions on private lands or within the wildland-urban interface areas. This study addressed gaps in existing wildfire risk assessments, that do not explicitly consider cross-boundary fire transmission and do not identify the sources of fire. The results can be used by State, Federal, and local fire planning organizations to help improve risk mitigation programs.

Climate change is expected to increase fire activity in many regions of the globe, but the relative role of human vs. lightning-caused ignitions on future fire regimes is unclear. We developed statistical models that account for the spatiotemporal ignition patterns by cause in the eastern slopes of the Cascades in Oregon, USA. Projected changes in energy release component from a suite of climate models were used with our model to quantify changes in frequency and extent of human and lightning-caused fires and record-breaking events based on sizes of individual fires between contemporary (2006 −2015) and mid-century conditions (2031–2060). No significant change was projected for the number of human-caused fire ignitions, but we projected a 14% reduction in lightning-caused ignitions under future conditions. Mean fire sizes were 31% and 22% larger under future conditions (2031–2060) for human and lightning-caused ignitions, respectively. All but one climate model projected increased frequency of record-breaking events relative to the contemporary period, with the largest future fires being about twice the size of those of the contemporary period. This work contributes to understanding the role of lightning- and human-caused fires on future fire regimes and can help inform successful adaptation strategies in this landscape.

Although nitrogen deposition and tropospheric ozone have impacted California forests for decades, broad scale studies of these impacts on forest growth and mortality are lacking. Because of the summer-dry climate over most of the state, forest responses to air pollution are expected to differ from more mesic climates. In this study, data from US Forest Service Forest Inventory and Analysis (FIA) permanent (remeasured) plots were combined with modelled atmospheric N and S deposition and an ozone exposure index to evaluate tree growth and mortality responses in California. Seven of 18 species exhibited significantly greater carbon increment (CI) in tree boles as N deposition increased, though the magnitude of the effect was quite small in most California forests. However, increases in CI were substantial in the coastal ecosections of central and northern California where precipitation and fog exposure are greatest. Redwood (Sequoia sempervirens (D. Don) Endl.) trees exhibited the strongest CI response to N deposition. Our model results imply a mean CI increase of 4.2 kg ha−1 yr−1 of C per kg ha−1 yr−1 of N deposition statewide versus 13.6 in the Central and Northern California Coast ecosections, where > 50% of the trees are redwood or tanoak (Lithocarpus densiflorus (Hook. & Arn.) Rehd.). Increased carbon sequestration rates in response to N deposition in these California coastal regions were similar to increases reported for Europe and global estimates. Nitrogen and S deposition significantly increased the odds of top damage and trees with crown damage exhibited higher mortality, although the effect was small. Elevated ozone exposure was associated with significantly larger rates of overall tree growth. However, for ozone-sensitive ponderosa pine at moderate ozone levels (ozone index values of ca. 20–30 ppb) and moderately-elevated N deposition (15–25 kg ha−1 hr−1), CI begins to decline, before increasing at higher pollution levels, presumably because of the fertilizing effect of N deposition; although data are limited for these more polluted conditions. Sulfur deposition in California forests was low, ranging from 0.3 to 3.1 kg ha−1 yr−1, but was associated with positive growth response in seven coniferous species. The combined effect of N and S deposition and ozone exposure statewide is a net increase in bole CI. However, aridity reduces the stimulatory growth effect of N deposition, and alters the threshold, capacity and sometimes the direction (e.g., S deposition) of the CI response to deposition, factors that need to be considered in global change models.

As the climate in California warms and wildfires become larger and more severe, satellite-based observational tools are frequently used for studying impact of those fires on air quality. However little objective work has been done to quantify the skill these satellite observations of smoke plumes have in predicting impacts to PM2.5 concentrations at ground level monitors, especially those monitors used to determine attainment values for air quality under the Clean Air Act. Using PM2.5 monitoring data from a suite of monitors throughout the Central California area, we found a significant, but weak relationship between satellite-observed smoke plumes and PM2.5 concentrations measured at the surface. However, when combined with an autoregressive statistical model that uses weather and seasonal factors to identify thresholds for flagging unusual events at these sites, we found that the presence of smoke plumes could reliably identify periods of wildfire influence with 95% accuracy.

Wildland fire suppression practices in the western United States are being widely scrutinized by policymakers and scientists as costs escalate and large fires increasingly affect social and ecological values. One potential solution is to change current fire suppression tactics to intentionally increase the area burned under conditions when risks are acceptable to managers and fires can be used to achieve long-term restoration goals in fire adapted forests. We conducted experiments with the Envision landscape model to simulate increased levels of wildfire over a 50-year period on a 1.2 million ha landscape in the eastern Cascades of Oregon, USA. We hypothesized that at some level of burned area fuels would limit the growth of new fires, and fire effects on the composition and structure of forests would eventually reduce future fire intensity and severity. We found that doubling current rates of wildfire resulted in detectable feedbacks in area burned and fire intensity. Area burned in a given simulation year was reduced about 18% per unit area burned in the prior five years averaged across all scenarios. The reduction in area burned was accompanied by substantially lower fire severity, and vegetation shifted to open forest and grass-shrub conditions at the expense of old growth habitat. Negative fire feedbacks were slightly moderated by longer-term positive feedbacks, in which the effect of prior area burned diminished during the simulation. We discuss trade-offs between managing fuels with wildfire versus prescribed fire and mechanical fuel treatments from a social and policy standpoint. The study provides a useful modeling framework to consider the potential value of fire feedbacks as part of overall land management strategies to build fire resilient landscapes and reduce wildfire risk to communities in the western U.S. The results are also relevant to prior climate-wildfire studies that did not consider fire feedbacks in projections of future wildfire activity.

We analyzed wildland fire occurrence and size data from Sardinia, Italy, and Corsica, France, to examine spatiotemporal patterns of fire occurrence in relation to weather, land use, anthropogenic features, and time of year. Fires on these islands are largely human caused and can be attributed to negligence, agro-pastoral land use, and arson. Of particular interest was the predictive value of a fire weather index (FWI) that is widely used by fire managers to alert suppression crews. We found that an increase in the FWI from 30 to 60 produced on average an approximate eightfold increase in the odds of a large fire, regardless of the time of year during the fire season or land use type. Total area burned per fire season was positively correlated with the number of days with FWI > 40 over the period studied. Strong interactions between time of year and land type were also observed for both the probability of ignition and large fire. For example, the estimated odds of a large fire on agricultural lands in southern Sardinia was approximately 10 times larger than the forest and shrubland land type for areas close to roads, early (May) in the fire season. Conversely, toward the end of the fire season (September), we estimated the odds of a large fire in these same areas at about half the value estimated for the forest land classes. Of the explanatory variables analyzed, only FWI had an effect on the probability of a large fire (P

Surface ozone (O3) was measured at the Devils Postpile National Monument (DEPO), eastern Sierra Nevada Mountains, California, during the 2007 (low-fire) and 2008 (high-fire) summer seasons. While mean and median values of O3 concentrations for the 2007 and 2008 summer seasons were similar, maximum O3 concentrations in June and August 2008 were higher than in any month of the 2007 summer season. This increase of maximum concentrations in the high-fire year is attributed to emissions of O3 precursors from wildland fires upwind of DEPO in addition to transport of polluted air from the California Central Valley (CCV) and the San Francisco Bay Area (SFBA). Analysis of backward trajectories for high O3 episodes in June 2007 and 2008 showed the lowest O3 pollution at DEPO when air masses originated over the Pacific Ocean (PO) and passed from West to East over PO and CCV at high altitudes. The highest O3 levels occurred when air masses originated in the vicinity of SFBA, swept through CCV in the NW–SE direction before reaching DEPO at low altitudes. Diurnal O3 concentrations were characterized by a sharp increase early morning and maximum values in late afternoon, followed by gradual evening decreases with very low pre-dawn minima, a phenomenon explained by local generation of O3 combined with the long range transport of polluted air masses from CCV, boundary layer destruction and surface deposition. Once in 2007, and 3 times in 2008, the 8-h average concentration exceeded 75 ppb, counting towards exceedance of the present primary Federal O3 standard (4th highest 8 h concentration

We use the simulation model Envision to analyze long-term wildfire dynamics and the effects of different fuel management scenarios in central Oregon, USA. We simulated a 50-year future where fuel management activities were increased by doubling and tripling the current area treated while retaining existing treatment strategies in terms of spatial distribution and treatment type. We modeled forest succession using a state-and-transition approach and simulated wildfires based on the contemporary fire regime of the region. We tested for the presence of temporal trends and overall differences in burned area among four fuel management scenarios. Results showed that when the forest was managed to reduce fuels it burned less: over the course of 50 years there was up to a 40% reduction in area burned. However, simulation outputs did not reveal the expected temporal trend, i.e., area burned did not decrease progressively with time, nor did the absence of management lead to its increase. These results can be explained as the consequence of an existing wildfire deficit and vegetation succession paths that led to closed canopy, and heavy fuels forest types that are unlikely to burn under average fire weather. Fire (and management) remained relatively rare disturbances and, given our assumptions, were unable to alter long-term vegetation patterns and consequently unable to alter long-term wildfire dynamics. Doubling and tripling current management targets were effective in the near term but not sustainable through time because of a scarcity of stands eligible to treat according to the modeled management constraints. These results provide new insights into the long-term dynamics between fuel management programs and wildfire and demonstrate that treatment prioritization strategies have limited effect on fire activity if they are too narrowly focused on particular forest conditions.

We used simulation modeling to analyze wildfire exposure to social and ecological values on a 0.6 million ha national forest in central Oregon, USA. We simulated 50,000 wildfires that replicated recent fire events in the area and generated detailed maps of burn probability (BP) and fire intensity distributions. We also recorded the ignition locations and size of each simulated fire and used these outputs to construct a fire source–sink ratio as the ratio of fire size to burn probability. Fire behavior was summarized for federal land management designations, including biological conservation reserves, recreational sites, managed forest, and wildland urban interface. Burn probability from the simulations ranged from 0.00001 to 0.026 within the study area (mean = 0.0023), and exhibited substantial variation among and within land designations. Simulated fire behavior was broadly related to gradients in fire regimes, although the combined effects of fuel, topography, and simulated weather resulted in fine scale patterns not reflected in ecological and vegetation data. Average BP for the northern spotted owl (Strix occidentalis caurina) nesting sites ranged from 0.0002 to 0.04. Among the 130 different wildland urban interface areas, average BP varied from 0.0001 to 0.02. Spatial variation in the source–sink ratio was pronounced, and strongly affected by the continuity and arrangement of surface and canopy fuel. We discuss the management implications in terms of prioritizing fuel management activities and designing conservation strategies on fire prone landscapes within the 177 million ha national forest network.

Large wildfire occurrence and burned area are modeled using hydroclimate and landsurface characteristics under a range of future climate and development scenarios. The range of uncertainty for future wildfire regimes is analyzed over two emissions pathways (the Special Report on Emissions Scenarios [SRES] A2 and B1 scenarios); three global climate models (Centre National de Recherches Météorologiques CM3, Geophysical Fluid Dynamics Laboratory CM2.1 and National Center for Atmospheric Research PCM1); three scenarios for future population growth and development footprint; and two thresholds for defining the wildland-urban interface relative to housing density. Results were assessed for three 30-year time periods centered on 2020, 2050, and 2085, relative to a 30-year reference period centered on 1975. Increases in wildfire burned area are anticipated for most scenarios, although the range of outcomes is large and increases with time. The increase in wildfire burned area associated with the higher emissions pathway (SRES A2) is substantial, with increases statewide ranging from 36% to 74% by 2085, and increases exceeding 100% in much of the forested areas of Northern California in every SRES A2 scenario by 2085. UCR::Vicerrectoría de Investigación::Unidades de Investigación::Ciencias Básicas::Centro de Investigaciones Geofísicas (CIGEFI)

Data from four continuous ozone and weather monitoring sites operated by the National Park Service in Sierra Nevada, California, are used to develop an ozone forecasting model and to estimate the contribution of wildland fires on ambient ozone levels. The analyses of weather and ozone data pointed to the transport of ozone precursors from the Central Valley as an important source of pollution in these National Parks. Comparisons of forecasted and observed values demonstrated that accurate forecasts of next-day hourly ozone levels may be achieved by using a time series model with historic averages, expected local weather and modeled PM values as explanatory variables. Results on fire smoke influence indicated occurrence of significant increases in average ozone levels with increasing fire activity. The overall effect on diurnal ozone values, however, was small when compared with the amount of variability attributed to sources other than fire.

From 1997 to 2006, we used mark-recapture models to estimate the site fidelity of 1250 Sierra Nevada yellow- legged frogs (Rana sierrae) in Kings Canyon National Park, California, USA, during their three main activity periods of overwintering, breeding, and feeding. To quantify site fidelity, the tendency to return to and reuse previously occupied habitats, we used multistrata models (with water bodies as the strata) and potential function analyses. The probability of returning to previously used water bodies during all activity periods was typically greater than 80% and always greater than the probability of moving to other water bodies. Site fidelity models (with lake-specific movement transitions) were favored over those models that held movement transitions equal between lakes. Potential function analyses demonstrated that frogs were most strongly attracted to their original capture lakes rather than moving to the nearest available breeding or overwintering lake. Under current disturbances in high-elevation Sierra Nevada lakes (exotic trout, climate change), site fidelity is problematic because frogs return to lakes subject to drying or those with fish rather than dispersing to other lakes. Future recovery of declining species will need to focus efforts towards restoring habitats when animals maintain strong site fidelity even when their habitats deteriorate. Resume´ : De 1997 a 2006, des modeles de marquage-recapture nous ont servi a estimer la fideliteau site de 1 250 gre- nouilles a pattes jaunes de la Sierra Nevada (Rana sierrae) dans le parc national de King's Canyon, Californie, E ´ .-U., du- rant leur trois principales periodes d'activite ´, soit l'hivernage, la reproduction et l'alimentation. Afin de quantifier la fideliteau site, c'est-a `-dire la tendance a retourner dans des habitats utilises anterieurement et ales exploiter de nouveau, nous utilisons des modeles astrates multiples (les plans d'eau representant les strates) et des analyses fonctionnelles poten- tielles. La probabilitede retourner dans un plan d'eau dejautiliseest, dans l'ensemble des periodes d'activite ´ ,g enerale- ment de plus de 80 % et toujours plus elevee que la probabilitede se deplacer vers un autre plan d'eau. Nous preferons les modeles de fideliteau site avec des transitions de deplacement specifiques achaque lac aux modeles qui presupposent que les transitions de deplacement sont egales pour tous les lacs. Les analyses fonctionnelles potentielles montrent que les gre- nouilles sont beaucoup plus fortement attirees vers le lac de leur premiere capture que portees ase deplacer vers le lac le plus proche de reproduction ou d'hivernage. Dans les conditions actuelles de perturbations dans les lacs de haute altitude de la Sierra Nevada (truites exotiques, changement climatique), la fideliteau site pose des problemes parce que les gre- nouilles peuvent retourner ades lacs sujets a la dessiccation ou des lacs contenant des poissons, plutot que se disperser vers d'autres lacs. La recuperation future des especes en declin devra concentrer ses efforts sur la restauration des habitats lorsque les animaux conservent une forte fideliteau site meme quand leurs habitats se deteriorent. (Traduit par la Redaction)

Combustion emissions and strong Santa Ana winds had pronounced effects on patterns and levels of ambient ozone (O3) in southern California during the extensive wildland fires of October 2007. These changes are described in detail for a rural receptor site, the Santa Margarita Ecological Reserve, located among large fires in San Diego and Orange counties. In addition, O3 changes are also described for several other air quality monitoring sites in the general area of the fires. During the first phase of the fires, strong, dry and hot northeasterly Santa Ana winds brought into the area clean continental air masses, which resulted in minimal diurnal O3 fluctuations and a 72-h average concentration of 36.8 ppb. During thesecond phase of the fires, without Santa Ana winds present and air filled with smoke, daytime O3 concentrations steadily increased and reached 95.2 ppb while the lowest nighttime levels returned to w0 ppb. During that period the 8-h daytime average O3 concentration reached 78.3 ppb, which exceeded the federal standard of 75 ppb. After six days of fires, O3 diurnal concentrations returned to pre-fire patterns and levels.

Foliage from loblolly (Pinus taeda L.), slash (P. elliottii Engelm.), and longleaf (P. palustris Mill.) pines was hand plucked to isolate the effects of level and timing of foliage removal on growth and mortality. Slash and loblolly pine received one of five defoliation treatments during one of four months: January, April, July, or October. Longleaf pine received one of three defoliation treatments during one of five 2-week periods from September to December. Growth and mortality were measured for each species for several years after treatment. Increased defoliation resulted in increased growth loss. Only total October defoliation caused significant mortality in loblolly and longleaf pine and moderate mortality in slash pine. Estimated difference in volume between control and defoliated trees after 5 or 6 years ranged from 0 to 2.3 ft3 (about 30 percent) per tree depending on species, defoliation season, and amount of defoliation. Trees receiving 95 percent or greater defoliation were generally smaller than the control trees after 13 or 19 years posttreatment.

The ability to forecast the number and location of large wildfire events (with specified confidence bounds) is important to fire managers attempting to allocate and distribute suppression efforts during severe fire seasons. This paper describes the development of a statistical model for assessing the forecasting skills of fire-danger predictors and producing 1-month-ahead wildfire-danger probabilities in the western United States. The method is based on logistic regression techniques with spline functions to accommodate nonlinear relationships between fire-danger predictors and probability of large fire events. Estimates were based on 25 yr of historic fire occurrence data (1980–2004). The model using the predictors monthly average temperature, and lagged Palmer drought severity index demonstrated significant improvement in forecasting skill over historic frequencies (persistence forecasts) of large fire events. The statistical models were particularly amenable to model evaluation and production of probability-based fire-danger maps with prespecified precisions. For example, during the 25 yr of the study for the month of July, an area greater than 400 ha burned in 3% of locations where the model forecast was low; 11% of locations where the forecast was moderate; and 76% of locations where the forecast was extreme. The statistical techniques may be used to assess the skill of forecast fire-danger indices developed at other temporal or spatial scales.

The purpose of this paper is to present a procedure for obtaining approximate maximum likelihood estimates for compound binary response models. The extra binomial variation is incorporated into the model by adding random effects to the fixed effects on the probit (or logit) scale. Numerical integration techniques are used to arrive at a solution of the likelihood equations. The paper also presents an illustrating numerical example based on a large toxicological data set. The computations are carried out within the GLIM statistical package.

As the climate in California warms and wildfires become larger and more severe, satellite-based observational tools are frequently used for studying impact of those fires on air quality. However little objective work has been done to quantify the skill these satellite observations of smoke plumes have in predicting impacts to PM2.5 concentrations at ground level monitors, especially those monitors used to determine attainment values for air quality under the Clean Air Act. Using PM2.5 monitoring data from a suite of monitors throughout the Central California area, we found a significant, but weak relationship between satellite-observed smoke plumes and PM2.5 concentrations measured at the surface. However, when combined with an autoregressive statistical model that uses weather and seasonal factors to identify thresholds for flagging unusual events at these sites, we found that the presence of smoke plumes could reliably identify periods of wildfire influence with 95% accuracy.

Chaparral watersheds associated with Mediterranean-type climate are distributed over five regions of the world. Because brushland soils are often shallow with low water holding capacities, and are on slopes prone to erosion, disturbances such as fire can adversely affect their physical properties. Fire can also increase the spatial coverage of soil water repellency, reducing infiltration, and, in turn, increasing overland flow and subsequent erosion. We studied the impacts of fire on soil properties by collecting data before and after a prescribed burn conducted during Spring 2001 on the San Dimas Experimental Forest, southern California. The fire removed the litter layer and destroyed the weak surface soil structure; leaving a thin band of ash and char on top of, and mixed in with, an unstable, granular soil of loose consistency. Median litter thickness and clay content were significantly decreased after fire while soil bulk density increased. At 7 d post-fire, soil surface repellency in the watershed was significantly higher than prior to the burn. At 76 d post-fire, surface soil water repellency was returning to near pre-fire values. At the 2 and 4 cm depths, 7 d post-fire soil repellency was also significantly higher than pre-fire, however, conditions at 76 d post-fire were similar to pre-fire values. Variability in soil water repellency between replicates within a given 15 × 15 cm site was as large as the variability seen between sites over the 1.28 ha watershed. The increase in post-fire persistence of water repellency was largest beneath ceanothus ( Ceanothus crassifolius ) as compared to a small increase beneath chamise ( Adenostoma fasciculatum ). However, pre-fire persistence was higher under chamise than for ceanothus. Post-fire changes to soil properties may increase the watershed hydrologic response, however the mosaic distribution of water repellency may lead to a less severe increase in hydrologic response than might be expected for a spatially more homogenous increase in repellency.

SUMMARY Forest fires are an important societal problem. They cause extensive damage and substantial funds are spent preparing for and fighting them. This work develops a stochastic model useful for probabilistic risk assessment, specifically to estimate chances of fires at a future time given explanatory variables. Questions of interest include: Are random effects needed in the risk model? and if yes, How is the analysis to be implemented? An exploratory data analysis approach is taken using both fixed and random effects models for data concerning the Federal Lands in the state of California during the period 2000–2003. Published in 2006 by John Wiley & Sons, Ltd.

SUMMARY We describe the use of bivariate stochastic differential equations (SDE) for modeling movements of 216 radiocollared female Rocky Mountain elk at the Starkey Experimental Forest and Range in northeastern Oregon. Spatially and temporally explicit vector fields were estimated using approximating difference equations and nonparametric regression techniques. Estimated vector fields of movement were mapped onto the project area at selected times of the day to examine spatial patterns of movement in relation to topography. Using the concept of a potential function, we were able to study the influence of roads and grassland foraging areas on elk movements. Doing so we identified broad spatial patterns of elk movements and showed the time dependent effects of habitat features within the habitat mosaic at Starkey. Our analyses quantify the cycles of movements in spring and summer in terms of attraction or repulsion to specific habitat features, and illustrate the magnitude, timing and direction of these movements. An extensive list of references is included. Published in 2004 by John Wiley & Sons, Ltd.

Terrestrial mollusks are important components of forest ecosystems, yet we know very little about the distribution and habitat of many of these species. We sampled for terrestrial mollusks in northern California with the goal of estimating the geographic ranges and developing predictive habitat models for five species that were assumed to be sensitive to land management activities. The species of interest were Ancotrema voyanum, Helminthoglypta talmadgei, Monadenia churchi, Monadenia fidelis klamathica, and M. f. ochromphalus. We randomly selected 308 plots for sampling from a grid of points across a 2.2 million-ha study area. We used Generalized Additive Models to estimate each mollusk's geographic range and to develop predictive habitat models within their ranges. Models were developed at one microscale (1 ha) and six mesoscales (ranging from 12.5 to 1250 ha) using vegetation, physical, climatic, and spatial location covariates. Estimated geographic ranges varied from 4770 to 15 795 km2. Predictive habi...

This work presents an exploratory data analysis of the trajectories of deer and elk moving about in the Starkey Experimental Forest and Range in eastern Oregon. The animals’ movements may be affected by habitat variables and the behavior of the other animals. In the work of this paper a stochastic differential equation-based model is developed in successive stages. Equations of motion are set down motivated by corresponding equations of physics. Functional parameters appearing in the equations are estimated nonparametrically and plots of vector fields of animal movements are prepared. Residuals are used to look for interactions amongst the movements of the animals. There are exploratory analyses of various sorts. Statistical inferences are based on Fourier transforms of the data, which are unequally spaced. The sections of the paper start with motivating quotes and aphorisms from the writings of John W. Tukey.

Human-caused fragmentation of habitats is threatening an increasing number of animal and plant species, making an understanding of the factors influencing patch occupancy ever more important. The overall goal of the current study was to develop probabilistic models of patch occupancy for the mountain yellow-legged frog (Rana mus- cosa). This once-common species has declined dramatically, at least in part as a result of habitat fragmentation resulting from the introduction of predatory fish. We first describe a model of frog patch occupancy developed using semiparametric logistic regression that is based on habitat characteristics, fish presence/absence, and a spatial location term (the latter to account for spatial autocorrelation in the data). This model had several limitations in- cluding being constrained in its use to only the study area. We therefore developed a more general model that incorporated spatial autocorrelation through the use of an autocovariate term that describes the degree of isolation from neighboring frog populations (autologistic model). After accounting for spatial autocorrelation in patch occupancy, both models in- dicated that the probability of frog presence was strongly influenced by lake depth, elevation, fish presence/absence, substrate characteristics, and the degree of lake isolation. Based on cross-validation procedures, both models provided good fits to the data, but the autologistic model was more useful in predicting patch occupancy by frogs. We conclude by describing a possible application of this model in assessing the likelihood of persistence by frog

The concern is with the properties of stochastic differential equations (SDEs) describing the motion of particles in 3 dimensional space, on the sphere or in the plane. There is consideration of the case where the drift function comes from a potential function. There is study of SDEs whose parameters are periodic in time. These are useful for incorporating circadian rhythm in the behavior. The cases of a seal in a frozen lake in Alaska, an elephant seal migrating a great distance in the Pacific Ocean and of a group of "free-ranging" elk in a reserve in Oregon are referred to. For the elk nonparametric estimates of the drift and variance terms of an SDE model are discussed and evaluated and the fit of the model assessed. One issue is how to include explanatories, beyond location and time, in the model. A number of questions motivated by the wildlife motion concerning diffusion processes of the type considered are posed at the end of the paper.

Topography, weather, and fuels are known factors driving fire behavior, but the degree to which each contributes to the spatial pattern of fire severity under different conditions remains poorly understood. The variability in severity within the boundaries of the 2006 wildfires that burned in the Klamath Mountains, northern California, along with data on burn conditions and new analytical tools, presented an opportunity to evaluate factors influencing fire severity under burning conditions representative of those where management of wildfire for resource benefit is most likely. Fire severity was estimated as the percent change in canopy cover (0–100%) classified from the Relativized differenced Normalized Burn Ratio (RdNBR), and spatial data layers were compiled to determine strength of associations with topography, weather, and variables directly or indirectly linked to fuels, such as vegetation type, number of previous fires, and time since last fire. Detailed fire progressions were used to estimate weather (e.g., temperature, relative humidity, temperature inversions, and solar radiation) at the time of burning. A generalized additive regression model with random effects and an additional spatial term to account for autocorrelation between adjacent locations was fitted to fire severity. In this fire year characterized by the relative absence of extreme fire weather, topographical complexity most strongly influenced severity. Upper- and mid-slopes tended to burn at higher fire severity than lower-slopes. East- and southeast-facing aspects tended to burn at higher severity than other aspects. Vegetation type and fire history were also important predictors of fire severity. Shrub vegetation was more likely to burn at higher severity than mixed hardwood/conifer or hardwood vegetation. As expected, fire severity was positively associated with time since previous fire, but the relationship was non-linear. Of the weather variables analyzed, temperature inversions, common in the complex topography of the Klamath Mountains, showed the strongest association with fire severity. Inversions trapped smoke and had a dampening effect on severity within the landscape underneath the inversion. Understanding the spatial controls on mixed-severity fires allows managers to better plan for future wildfires and aide in the decision making when managing lightning ignitions for resource benefit might be appropriate.

We used analyses based on surveys of > 1700 water bodies in a 100,000-ha area in the John Muir Wilderness (JMW) and Kings Canyon National Park (KCNP) to determine the influence of nonnative trout on the distribution and abundance of Hyla regilla in the High Sierra Nevada. At the landscape scale (JMW compared to KCNP), a negative relationship between trout and frogs in lakes was evident. In the JMW study area where trout are more abundant, only 7.2% of all water bodies contained H. regilla versus 26.6% in the KCNP study area. Also, the percentage of the total water body surface area containing H. regilla was 19.4 times higher in the KCNP study area than in the JMW study area. Hyla regilla were most abundant in portions of KCNP where the probability of finding lakes with trout is lowest and least abundant in the northern part of the JMW where the probability of finding lakes with trout is highest. At the water body scale, after accounting for the effects of all significant habitat and isolation var...

Prescribed fire is a management tool used by wildland resource management organizations in many ecosystems to reduce hazardous fuels and to achieve a host of other objectives. To study the effects of fire in naturally accumulating fuel conditions, the ambient soil temperature is monitored beneath prescribed burns. In this study we developed a stochastic model for temperature profiles (values at 15 minute intervals) recorded at four depths beneath the soil during a large prescribed burn study. The model was used to assess the temporal fit of the data to particular solutions of the heat equation. We used a random effects model to assess the effects of observed site characteristics on maximum temperatures and to estimate risks of temperatures exceeding critical levels in future similar prescribed fires. Contour plots of estimated risks of temperatures exceeding 60°C for a range of fuel levels and soil depths indicated high risks of occurrence, especially when the moisture levels are low. However, the natural variability among sites seems to be large, even after controlling fuel and moisture levels, resulting in large standard errors of predicted risks.

It is widely believed that stream salmonids select spawning sites based on water depth, water velocity, and substrate size. Attempts to predict spawning locations using these habitat features have met with little success, however. In this study, we used nonparametric logistic regression to determine what habitat features were associated with the locations chosen by spawning California golden trout (Oncorhynchus mykiss aguabonita). From this nonparametric model, we developed a parametric model that incorporated the habitat features most strongly associated with spawning sites and used this model to calculate the probability of use by spawning golden trout for specific stream locations. The overall nonparametric model was highly significant and explained 62% of the variation in spawning location. Four of the eight habitat variables, substrate size, water depth, water velocity, and stream width, had highly significant effects and alone explained 59% of the variation in spawning location. The results of a cross-validation procedure indicated that the parametric model generally provided a good fit to the data. These results indicate that location-specific probabilities of use were predictable based on easily measured habitat characteristics and that nonparametric regression, an approach still rarely used in ecological studies, may have considerable utility in the development of fish-habitat models. Given the escalating pace at which fish habitats are being altered, such models are increasingly important in predicting the effects of these alterations on populations.

Analysis of bioassays data from experiments with mixtures require use of special statistical techniques, especially when responses are recorded over time. Often, response curves from bioassays with mixtures do not fit a probit or logit curve. In addition, observations recorded over time are not independent. As a result, standard statistical packages for probit or logit analysis are not applicable. In these bioassay s. the shape of the response curve depends on the type of interactions between mixture components. Here we present models that describe response (e.g., mortality) curves over time when mixtures are applied to insects. The shapes of the response curves depend on the modes of both action and interaction between the mixture compounds. We describe a procedure to assess whether the assumed modes of interaction (e.g., independent action or similar action additive models) are appropriate. We demonstrate use of the statistical methods with data obtained when a baculovirus and a synthetic organic insecticide were mixed.

Our investigation estimates time–temperature relationships and demonstrates the effect of rapid heating on mortality of Caribbean fruit fly, Anastrepha suspensa (Loew). Mature larvae (third instars) in water were exposed to each of seven temperatures (44, 45, 46, 47, 48, 49, and 50 °C) and each of five power levels (11, 27.5, 55, 88, and 122 W) in a research-quality microwave oven. Controls were immersed in water for 30 min and not exposed to microwave energy. Data were analyzed by a probit model with three explanatory variables. The variables were time to reach target temperature, power, and final temperature. Temperature needed to control the larvae increased as power increased. Of the power–temperature levels, the only combinations that resulted in > 99% mortality were the lowest power (16 W) at 49 or 50 °C and 30 W at 50 °C. Time required for > 99% mortality decreased with increased power. Thus, as power delivered to larvae increased and time needed to reach exposure temperature deceased, percent mortality decreased. We conclude that rapid heating imposes serious constraints on the use of heat-induced mortality; this result raises important questions that must be addressed because quarantine security may be jeopardized.

The fall rate of nearly 600 lodgepole pines(Pinus contorta)killed by the mountain pine beetle (Dendroctonus ponderosae) in central Oregon was investigated in thinned and unthinned stands. Estimates were obtained by fitting a complementary log-log model to the conditional probabilities of trees falling within a given year. Snags began falling 3 yr after death in thinned stands and 5 yr in unthinned stands. Small trees fell slightly faster than large trees in thinned stands, but tree size was not a factor in the fall rate in unthinned stands. In thinned stands, 50% were down in 8 yr and 90% were down in 12 yr. In unthinned stands, 50% were down in 9 yr and 90% were down in 14 yr. No particular calendar year had tree fall that was significantly greater than average. All beetle-killed trees broke off at the ground when they fell. The rate that trees fall in different environments may be related to the speed of bole decay at the ground level. West. J. Appl. For. 13(1):23-26.

Can fire potential forecasts assist with pre-positioning of fire suppression resources, which could result in a cost savings to the United States government? Here, we present a preliminary assessment of the 7-Day Fire Potential Outlook forecasts made by the Predictive Services program. We utilized historical fire occurrence data and archived forecasts to assess how well the 7-Day Outlook predicts wildfire ignitions and escaped fires, ultimately to help characterize the effectiveness of this tool for prepositioning national firefighting resources. The historical fire occurrence data track ignitions on all land ownerships; from this dataset, we established number and location of ignitions and final fire size for the years 2009-2011 for Predictive Service Areas (PSAs) within the Northwest and Southwest Geographic Areas. These data were then matched to the corresponding PSA and appropriate forecast for each of the seven days prior to the ignition date. Final fire size was used as a metric to establish whether an ignition escaped initial attack, with fires greater than 121.4 hectares (300 acres) considered escaped. Our results show that 7-Day Outlook values yield better-than-random prediction of large fire activity, although there is wide variation in this relationship among individual PSAs. In addition, the number of escaped fires increased with the number of ignitions, with this relationship showing a distinct regional pattern. Fires were more likely to escape during certain times of the year, with this season being earlier in the Southwest than in the Northwest. Significantly higher numbers of escaped fires per ignition occurred during days considered to be high risk by the meteorologist than on lower-risk days.

Bioassays with MVP, a genetically engineered form of a toxin of Bacillus thuringiensis subsp. kurstaki, demonstrated that a plateau in concentration response depended on the temperature and time when mortality was assessed. The presence of a plateau did not indicate resistance in diamondback moth, Plutella xylostella (L.). Comparisons of LC50S at 25 degrees C indicated that MVP acted more slowly than a conventional B. thuringiensis subsp. kurstaki formulation (Dipel 2X). However, LT50S of the preparations at a given concentration did not vary substantially. At 28 degrees C, responses were approximately equal over time. To avoid artifacts in analyses of concentration-response data, we recommend that bioassays be designed carefully to account for subtle differences in the behavior of genetically engineered versus conventional preparations of B. thuringiensis.

Our investigation of responses of Colorado potato beetle, Leptinotarsa decemlineata (Say), to Bacillus thuringiensis subsp. tenebrionis over 83 wk; diamondback moth, Plutella xylostello (L.), to B. thuringiensis subsp. kurstaki (cloned into Pseudomonas florescens ) over 37 wk; and western spruce budworm, Choristoneura occidentalis Freeman, to pyrethrins and DDT over 91 generations indicated that ratios at LC50, LC90, and LC99 varied among pesticides tested on the same species and among insect species tested with the same pesticide. Frequencies with which LCs were significantly different (based on bracketing of the value 1.0 by 95% CI of each ratio) compared with the standard (the lowest LC in the data set) were extremely high (>95%) in all tests except those with pyrethrins on western spruce budworm. For example, toxicity ratios for Colorado potato beetle larvae fed B. thuringiensis subsp. tenebrionis were as high as 12.8 at LC50, 37.0 at LC90, and 155 at LC99 over 83 wk; the maximum upper 95% CL for a ratio were 18.6, 168, and 1,000, respectively, at these LCs. We conclude that the extent of natural variation must be investigated before biologically important changes can be identified with any certainty. The conventional practice of using ratios of one LC to another in studies to detect resistance and other biological changes may lead to erroneous conclusions if natural variation in cohorts of a population and subsequent generations of the same genetic strain are unknown.