Introduction
Dr. Craig Allen, a USGS research ecologist with the Midcontinent Ecological Science Center, is speaking of the New Mexico forest ecosystems he knows best, but his words apply equally well to most of western North America. "If you're trying to understand past and present patterns on the landscape," Allen says, "first of all you need to know something about fire."
From the northern Rocky Mountains to the Southwest borderlands, wildland fires have burned and rejuvenated western forests over the course of millennia. And forests are not the only environments affected by fire; to a greater or lesser degree, fire influences the structure and dynamics of nearly all of the West's terrestrial ecosystems. In some, such as the chaparral brushlands of California, fire has been a strong force guiding the evolution of local plant life, and a constant regulator of ecological communities. In many desert habitats, on the other hand, fires have been far less frequent, but represent a more severe disturbance when they do occur.
Scientists and managers increasingly recognize the importance of fire as a natural component of ecological systems. But while fire is often a beneficial process, it is always, in the short term, a destructive one. The presence of fire has usually been seen as incompatible with both human land-use practices and aesthetics, and for over a century fires have been actively suppressed throughout the West.
The negative consequences of forest fire suppression can now be clearly seen. In many areas, disruption of the natural fire regime has produced overcrowded forests with vast accumulations of dry fuel. Blazes that break out under these conditions may be far more destructive than the normal fires of centuries past and are often extremely difficult or impossible to control. The absence of a regular fire cycle has also harmed many plant and animal species whose life histories are tightly linked to fire disturbance.
Across regions and among different forest types, the historical role of fire and the effects of recent fire suppression vary. And while fire suppression has fundamentally altered many forest ecosystems, the opposite is often true in grassland, shrubland and desert habitats. In these systems, fire incidence has been increasing, often due to the spread of non-native vegetation, with negative consequences for native plants and animals. Thus no single prescription for fire management will work in all areas. Programs of prescribed burning, highly successful in some forests, may not succeed in other habitats.
The challenge for managers seeking to restore more normal fire dynamics to a particular region is indeed, as Allen observes, to know something about fire: how fire has historically affected the local system, and how it functions today. Across the West, USGS researchers, in collaboration with scientists from numerous other agencies and institutions, are providing this information through detailed studies of fire history and fire ecology in different environments. USGS Director Charles G. "Chip" Groat says that, "Knowledge from these studies is forming the basis for new policies aimed at restoring fire cycles that will present a lower risk to human life and property, and help safeguard the stability and diversity of western ecosystems."
Sierra Nevada Forests
Since the 1960s, pioneering studies on the effects of both forest fires and decades of fire suppression have been carried out in the Sierra Nevada mountains of California in Yosemite, Sequoia and Kings Canyon National Parks. As in the Southwest, fire suppression in the Sierra Nevada has now led to conditions in which catastrophic fires may threaten the forests themselves. Suppression of lightning-caused fires has resulted in denser forests, invasion of open areas by trees and shrubs and large accumulations of woody debris.
Scientists and managers in the Sierra Nevada parks have long recognized the essential nature of fire in these forests and have responded over the years with an increasingly sophisticated fire restoration program using both prescribed burns and natural fires. At Yosemite, USGS fire ecologist Dr. Jan van Wagtendonk has devoted over a quarter-century of research to understanding what controls the behavior of forest fires, and how natural and prescribed fires can best be managed to reduce understory fuel loads and restore normal ecosystem dynamics.
Simple in overall conception, the use of fire in ecological restoration is a highly complex undertaking. Van Wagtendonk says that to be successful, fire management programs require a clear set of goals based on a detailed understanding of the role fire has played in the local forest environment. Managers also need extensive information regarding fuel loads, weather, topography and other factors to make informed decisions on where, when, how often and how hot to burn. "To know whether or not to allow a lightning fire to burn, managers need to know where it might spread in the next three months -- or the next three hours," van Wagtendonk says.
His current work has centered on the development of a new, high-resolution fuels map for Yosemite National Park. The map is based on satellite images of vegetative cover broken down into 30 by 30 meter squares, each representing one of 30 unique fuel categories. Additional data are provided by geographic information system (GIS) maps, aerial photographs and field measurements from more than 1,000 sites. All of this information is coupled with a computer model for predicting exactly where and how fast a given fire may spread.
The final product is a highly versatile tool for understanding fire behavior. Because of its relatively fine scale, van Wagtendonk says, the map captures the mosaic-like nature of surface fuels over fairly small areas. Studies have demonstrated that fire spread is highly sensitive to this kind of local variability in fuel type, but previous fuel maps derived from remote sensing data have been unable to capture this level of detail. Moreover, the depth of information contained in the map allows researchers to conduct both long-term and real-time predictive modeling.
The map and model have already been used on several occasions to predict the behavior of natural fires. From each such application, further refinements are made. In these initial tests, such as during Yosemite's Horizon Fire in 1994, the model performed well, said van Wagtendonk, providing managers with maps showing where fire perimeters would be at various future times, based on existing or changing weather conditions. The model has since been used to plan and execute prescribed burns in the park and to predict fire behavior on landscapes subjected to different techniques of understory fuel reduction, from mechanical thinning of trees to prescribed burning.
Van Wagtendonk says potential applications go beyond managing fires within the park. The mapping and data analysis techniques he has developed can in principle be extended to much larger areas, such as the entire Sierra Nevada. The fuels modeling package can also be used as a research tool. For example, scientists can approximate what the local landscape might look like without a history of fire suppression, by allowing past suppressed fires to "burn" and run their course on computers.
While advanced imaging and computer technology can help predict fire behavior in the future, tree ring analysis reveals fire patterns of centuries past. At Sequoia and Kings Canyon National Parks, USGS researchers and collaborators from the University of Arizona's Laboratory of Tree-Ring Research have put together the longest and most detailed fire histories anywhere. The records, assembled from fire scars in the annual growth rings of giant sequoias, extend back over 2,000 years, and show that fire typically burned on the floor of sequoia groves every 3 to 8 years.
USGS ecologist Dr. Nate Stephenson, from the Western Ecological Research Center, says the record shows how sequoias have responded to what has been, on a scale of centuries, an ever-changing climate and fire regime. The historical record shows a shifting matrix of low to moderate-intensity fires, with occasional hot spots of severe fire that open gaps in the forest and clear the way for sequoia regeneration. "The hot spots reduce competition so that the sequoia seedlings have chance," Stephenson says. Sequoia seeds require contact with bare soil in order to germinate, and this is possible only when fire has cleared away the layers of leaf litter and debris.
The loss of fire in sequoia groves has greatly affected the population. "Fire exclusion by humans has done more than the last three millennia of climate and fire regime changes," Stephenson says. "Essentially, when you exclude fire, sequoia reproduction crashes to zero." That means that in sequoia groves today, even the youngest trees are over a century old. Most areas in most groves have not burned for 100-130 years.
The good news, says Stephenson, is that the research message is reaching managers. He has studied the effects of different forest restoration measures including prescribed burning and mechanical thinning of trees. Unlike some other forest systems, Stephenson says, sequoia groves respond extremely well to prescribed burning alone, with no other treatment needed. "Where we have had prescribed fires, there's now a lot of sequoia reproduction -- enough that if it is maintained over the long term it will maintain the populations."
Is a burned forest a healthier forest? Certainly by some standards, but Stephenson prefers to say that fire restores stability and resiliency to forest ecosystems. "We're restoring a forest structure that's more stable, meaning if you give it a shove it's less likely to be bent out of shape. And it's more resilient, because if you do bend it out of shape it will bounce back quicker."
Fire management and restoration programs in the Sierra National Parks now reflect much of what researchers like van Wagtendonk and Stephenson have learned about the behavior and ecology of wildfires. The current prescribed burning program, says Stephenson, is highly successful. "It's an excellent example of how research has fed into management and changed management direction."
Nevertheless, says van Wagtendonk, "so much needs to be done, it's hard to get ahead of the game." One major constraint is smoke, which limits the amount of prescribed burning that can be done. Fire managers must work to stay within the bounds of clear air standards, and limit the amount of smoke descending on local communities. Stephenson says that while only a few prescribed fires create a smoke problem, these can erode public support for fire restoration. Continuing education is vital, he says, for people to understand that without some fire, both forests and human communities face the ever-growing danger of a major conflagration.
Interior and Coastal Shrublands
While the decline of old-growth forests has long been a high-profile issue in the West, the widespread loss of arid shrublands has gone practically unnoticed. But in the sagebrush ecosystems of the Great Basin and the Columbia River Basin, fire and a non-native plant species known as cheatgrass are together transforming ecological communities across a vast area. These changes may be irreversible, says USGS ecologist Dr. Steve Knick of the USGS Forest and Rangeland Ecosystem Science Center.
Knick studies these transformations at the Snake River Birds of Prey National Conservation Area in southwestern Idaho. Here, as in much of the Great Basin, the dominant vegetation -- sagebrush and other shrubs adapted to the harsh seasonal climate -- is disappearing. Of the roughly 100,000 hectares of shrubland present in the National Conservation Area in 1979, only 46,000 hectares remain.
To put it simply, Knick says, the shrubland is burning up. Wildfire incidence has increased by a factor of three since 1980, and fires are getting larger. But behind this increase -- and in turn capitalizing on it -- is the fast-spreading, exotic annual grass. "It's a synergistic thing," says Knick. "Cheatgrass promotes fire spread, and the larger fires eliminate more shrubs."
Knick says that fire has always been a factor in sagebrush ecosystems, creating openings in the shrub canopy and constraining the density of woody plants much the same as in forests. But an understory of native bunch grasses, which grow in isolated patches, tends to limit the intensity of blazes in these systems and prevent them from spreading over a wide area.
Cheatgrass, which has been advancing since the early 1900s, in part due to overgrazing and drought, creates a continuous carpet of fuel. And cheatgrass thrives on recently burned land, thus perpetuating the altered fire regime. If the shrubs in an area don't have time to recover before the next fire hits, they eventually disappear. "Fire has gone from maintaining a shrubland, to destroying a shrubland, to ultimately maintaining an exotic grassland," Knick says.
Using data from a number of sources including satellite imagery, historical records of fire frequency and behavior, and ground measurements of vegetation, Knick's team has developed a computer model for predicting long-term changes resulting from different scenarios of burning and regeneration of vegetation. The model shows that in shrublands with a cheatgrass understory, fire can easily trigger a rapid transition to grassland. But once established, these grassland systems tend to be relatively stable, even when fire is suppressed.
"The daunting thing is that it's going to take a long time to replace what has been destroyed in the last 20 years," Knick says. "We're looking at centuries if we rely only on natural processes for recovery." Knick's work suggests that preserving intact shrublands through more active fire suppression may be the only way to halt the losses. Restoration practices based on prescribed burning, as have been successfully carried out in forest ecosystems, may not work well in invaded shrublands.
"In forests you can use prescribed burning to remove a lot of the fine fuels, with the expectation that they are going to take several years to grow back," Knick says. "In shrublands dominated by cheatgrass, the cheatgrass will be back next year. It's using a disturbance to try to eliminate a species that likes disturbance."
While interior shrubland ecosystems have only a limited tolerance for fire, a very different kind of fire dynamic exists in the chaparral shrublands of coastal California. Ecologists have long known that chaparral ecosystems burn extensively and often, and much of the dominant vegetation in these systems is highly adapted to a fire-prone environment. Many plants have seeds that require fire to germinate, or need the kind of disturbed habitat fires leave behind in order to grow.
Dr. Jon Keeley, a USGS research ecologist with the Western Ecological Research Center, has studied the physiological adaptations that link the life cycles of chaparral vegetation with the natural regime of frequent brushfires. Upon reproduction, many species drop seeds that remain dormant in the soil "seed bank" until fire creates favorable growth conditions. When the area burns, these seeds receive a number of cues that may cause them to germinate. While seed germination in some species is stimulated by heat, in many others the onset of plant growth requires chemical exposure to combustion products such as charred wood.
Recently Keeley and Dr. C. J. Fotheringham, of California State University, Los Angeles, published a study demonstrating that for many species, smoke can also trigger seed germination. In some species smoke alone is sufficient to induce growth, while in others a combination of factors is required. Because of the diverse cues through which vegetation may respond to fire, blazes of different intensities or degrees of smoke production may result in different plants dominating the post-fire recovery. Of particular interest is their discovery, detailed last year in the journal Science, that nitrogen oxides, which are also important components of air pollution, are the chemicals in smoke responsible for germination of some species.
Keeley and his collaborators have also examined historical patterns of California shrubland wildfires. Life and property losses from shrubland fires in California have been increasing in recent decades. It has long been thought that fire suppression has played the same role in chaparral shrubland as it has in forests, creating a build-up of fuels that eventually leads to more destructive fires.
Surprisingly however, a close analysis of state fire records revealed a different story. This June, in the journal Science, Keeley and his co-authors reported that since 1910, chaparral fire frequency has not changed and fire size has not increased. The researchers found that large, intense fires were equally common in the years before widespread fire suppression as today, and do not appear to be the result of fuels build-up. In this highly fire-prone ecosystem, suppression efforts appear not to have greatly altered normal patterns of fire incidence. Keeley says the greater financial cost of fires today is more likely the result of constant urban expansion into areas subject to frequent burning.
The Mojave and Sonoran Deserts
In the Mojave Desert of California and Nevada, and the Sonoran Desert of Arizona, researchers are grappling with a fire and invasive species problem similar to that affecting Great Basin shrublands. Fire has not traditionally played a large role in organizing biological communities in these environments, where extremely arid conditions limit the density of vegetation. But in the deserts, too, alien grasses are now spreading, bringing rapid fire cycles with them.
Todd Esque, an ecologist with the USGS Western Ecological Research Center, says that although cheatgrass is not a big problem in the Mojave, other exotic species such as red brome are spreading fire through native communities that often have few evolved defenses against such disturbances. Esque and others are conducting detailed studies to better understand how increased fire size and frequency can affect desert ecosystems, and how native plants and animals can be protected. "We're trying to take a holistic view of the fire-weed cycle," says Esque. "We're looking at how fire changes nutrients in the soil, which changes the plants that are there, and in turn how animals respond to this dramatic change in habitat."
One of Esque's USGS collaborators, Dr. Matt Brooks, has studied recent historical changes in fire incidence and fire effects in the Mojave. Brooks says that while it is difficult to reconstruct long-term fire histories in desert systems, records from federal land management agencies do show an increase in Mojave Desert fires over the past two decades. Expanding human use of desert lands may be behind some of the increase, but Brooks says the pattern holds even in remote areas where fires are almost all lightning-caused.
"The increase in fires seems to be due to the alien annual grasses," Brooks says. These grasses often build up during years of heavy rainfall and, unlike many native annuals, their dry stalks may remain rooted in the ground for many years after they die, providing a lasting fuel source.
Brooks and others have also found that native Mojave Desert plants are often particularly vulnerable to fire. Although some species do resprout after burning if the fire intensity is not too high, few can tolerate successive burns. "If a second fire occurs before fire-damaged individuals have a chance to build back their above-ground, photosynthesizing biomass, they often die," says Brooks. "The grass-fire cycle reduces the return interval between fires to the point where most native desert shrubs and bunchgrasses cannot survive."
Much of Esque's own work is focused on understanding the mechanisms of invasion. He is carrying out a set of manipulative experiments comparing deliberately burned and unburned plots. Previous researchers have found that in desert environments, seed-eating rodents and ants often play a large role in determining the structure and composition of the vegetation. By removing ants and rodents from some of his plots, Esque can assess how the presence or absence of seed-eaters, along with changes in the surface vegetation and soil nutrients caused by fire, interact to determine the course of weed invasion and habitat transformation.
Esque also takes advantage of the "natural experiments" provided by desert wildfires. In 1994, for example, he began a collaborative studies with USGS research ecologist Dr. Cecil Schwalbe, of the Western Ecological Research Center, on the effects through time of a large fire in Saguaro National Park in Arizona. In the aftermath of the fire, the researchers quickly assembled a field team and began a detailed census and monitoring effort both inside the burn area and on adjacent unburned lands.
In documenting the ensuing changes to the area's plant and animal life, the study has focused on two of the most representative species of the Sonoran Desert: the saguaro cactus and the desert tortoise. Both suffered high mortality, and damaged saguaros continued to die several years after the fire -- which Schwalbe notes was of only moderate intensity. "Both tortoises and saguaros are long-lived species, which need very low annual mortality rates in order to maintain stable populations," Schwalbe says. "This fire resulted in a catastrophic loss for both of those species."
As in the Mojave, the fire problem in the Sonoran Desert is worsening. The 1994 fire in Saguaro National Park was spread by red brome. And Esque says his team's surveys in remote, unburned areas of the park have revealed that penetration by exotic grasses -- including a perennial, drought-adapted species from Africa known as buffelgrass -- is far worse than was previously known. "There wasn't a fire problem in this area before the exotic species came in," says Schwalbe. "Now we're seeing a biome conversion, from palo verde and saguaro habitat to a mesquite-acacia savannah with a Mediterranean exotic grass understory. That's the future of the Sonoran Desert -- especially near roads."
Southwest Forests
Some of the most extensive and detailed records of past fire activity come from the southwestern United States. Over thousands of years, this region's widespread ponderosa pine forests have been shaped and structured by fire. Historically, frequent low-intensity ground fires maintained open, park-like forests with grassy understories. Although such fires are often very local in nature, a broad historical perspective reveals regional-scale patterns of fire incidence and intensity, driven by climatic variability.
Dr. Craig Allen has studied the history and effects of fire in the Jemez Mountains of northern New Mexico since 1986. He and his collaborators employ several different methods for reconstructing the fire history of the Jemez and neighboring Sangre de Cristo mountains. Fires that do not kill a tree often leave a scar, which is recorded in the tree's annual growth ring. By carefully examining the tree rings, researchers can determine the year and often even the season in which the fire occurred.
Allen's team has put together over 4,500 fire dates, from over 600 trees, logs and stumps. "The Jemez is one of the better-sampled landscapes of its size anywhere," says Allen. "Very few areas have as much fire history." Analysis of tree rings is carried out in cooperation with Dr. Thomas Swetnam of the Laboratory of Tree Ring Research at the University of Arizona. The Jemez data form part of the lab's regional tree ring network for the entire Southwest, which contains regional fire history and climatological records for over 1,000 years.
Now Allen is using a different method to extend the Jemez fire history record back even further. In collaboration with Dr. Scott Anderson of Northern Arizona University, he is reviewing pollen and charcoal deposits in soil cores extracted from several northern New Mexico bogs. The cores contain a record of sediment deposition going back over 10,000 years. In one Jemez sample that has already been analyzed, Allen says, the contrast between the current century of fire suppression and the millennia that preceded it are clearly visible. "There is abundant charcoal throughout the core, except for the last couple of centimeters, corresponding to the past 125 years," he says. "It shows that fire has been an important ecological process here in the Jemez for at least 8,000 years."
Although humans have long shaped their landscapes through deliberate use of fire, Allen says fire patterns in the Southwest have largely been driven by the region's weather patterns. "Human ignitions were probably less important here than in most places on the planet," he says. Recent data show that the Jemez Mountains average about 16,000 lightning strikes per year, and Allen's analysis of fire suppression records for roughly 5,000 fires since 1909 indicate about 75 percent were of lightning origin.
Fire scars indicate that historically, blazes were most frequent in the dry spring and early summer period, before the arrival of the late-summer monsoon rains. Most burned only along the ground, clearing away debris and maintaining open, montane grasslands over large areas. Some of the trees Allen has sampled experienced more than 30 fires over the course of a few hundred years, without being killed.
Swetnam and climate change scientist Dr. Julio Betancourt, of the USGS Desert Laboratory, have shown that patterns of fire incidence in Allen's Jemez data are often mirrored across the broader Southwest region. The episodic occurrence of "regional fire years" appears to be associated with El Nio and La Nia events. Much of the Southwest is strongly affected by the weather patterns that characteristically follow these shifts in equatorial Pacific Ocean currents. El Nio years bring above-normal precipitation to the region, while La Nia years -- which often follow on the heels of El Nios -- are dry.
As might be expected, fire activity historically is greatest during La Nia events and droughts. But wet El Nio episodes play a role as well. In an environment in which water is often a limiting resource, wet years result in a rapid build-up of herbaceous understory vegetation. Years of intense regional fire activity often occur at the end of an El Nio-La Nia cycle, when this extra plant growth becomes a blanket of dry fuel across southwestern mountain ranges.
In the late 19th century, however, other factors came to dominate the region's fire regime. Allen says that in the arid Southwest, grazing has played at least as big a role as fire suppression in altering the natural pattern of frequent, low-intensity burns. Beginning in the 1880s, large numbers of cattle and sheep were introduced into southwestern forests. As grazers consumed the grasses and other herbaceous vegetation fires need in order to spread, fire activity dropped off. In addition, said Allen, the trails created by livestock over time probably constrained the spread of fire as well by breaking up the continuity of the surface fuels.
"The initial cessation of fires preceded active fire suppression by several decades," Allen says. As grasses were reduced and fires ceased to spread across the landscape, more trees were able to get established. The effects of grazing were then multiplied when fire suppression became the norm early this century. Eventually, open areas were replaced by dense tree stands. Allen says that historically, a typical density of ponderosa pines in the Southwest was around 100 stems per acre. Today, densities at many sites exceed 2,000 stems per acre.
In these dense forests, destructive insect outbreaks are common. And fires, when they do get established, now often leave the ground and climb "ladder fuels" into the treetops. "We did not start to see extensive crown fires in ponderosa pine forests until around the 1950s," Allen says. "It took that long for the forests to get dense enough and for the fuel conditions to change."
Tree ring records show that, in addition to the El Nio-La Nia cycles, periodic droughts and wet periods of much longer duration are also part of the normal climatic variability in the Southwest. A prolonged drought during the 1950s contributed to outbreaks of large, destructive fires at that time. However since then -- particularly over the last 20 years -- precipitation totals across most of the Southwest have been abnormally high -- a fact that Allen says should be cause for concern.
"It's been extremely good for tree growth, and a lot of extra forest biomass has accumulated on the landscape," he says. "The next time we have a significant drought of any sort, we can expect some very severe fire behavior. It's the scale and associated ecological effects of potential crown fires that we worry about. The forests across whole mountainsides can just go up, burning entire watersheds and resulting in severe post-fire erosion and flooding. Once such crown fires are in progress, we can't stop them through direct suppression methods."