Are big, severe wildfires normal?


The conventional wildfire wisdom goes something like this: Western forests are out of whack due to past fire suppression and logging practices. Forests that used to be open and free of undergrowth have turned into dense “dog hair” thickets of young trees that burn like kindling. Combine that with millions of acres of trees killed in a climate-change-fueled beetle outbreak, and we’ve ushered in an era of mega fires that are bigger, hotter and more intense than ever before.

We’ve heard this narrative over and over and over, and with good reason: there are a lot of scientific studies supporting it.

But a team of researchers from the University of Wyoming complicate that view with their latest study, which suggests that today’s wildfires aren’t any worse than those that burned in the late 19th century.

Bill Baker, a professor of ecology and geography, and Mark Williams, a recent graduate student, set out to test assumptions that dry, Western forests were historically more open and park like, and severe fires less frequent.

They set their sights on eastern Oregon, along Colorado’s Front Range, and on Arizona’s Black Mesa and Mogollon Plateau. First, they had to figure out what forests in these places looked like before European American settlement. They decided to combine tree-ring data, which has been widely used to reconstruct historic forests, with General Land Office surveys from the late 1800s. They recruited teams of students to browse through 13,000 hand-written records created by surveyors who were charting undeveloped federal land so it could be divided up and sold. The surveyors took meticulous notes on the landscape they walked through, putting bearing marks on trees every mile and describing when they entered a clearing, a burned forest or a stand of old-growth trees.

What the researchers found surprised them. Forests in the four regions were historically much denser than expected, especially in the Front Range. Open park-like forests made up less than half of all the forested areas they examined.

Baker and Williams also found that low-severity fires, which kill less than 20 percent of trees, weren’t as widespread as predicted. Instead, high-severity fires, which destroy more than 70 percent of trees in a burn area, were common in all four forests. Just as common, in fact, as they are today. The Front Range in particular, scorched by two destructive fires already this season, has a long history of big, severe fires.

The team from Wyoming rounded out their studies by criticizing common forest management practices. Except where necessary to protect homes, they believe widespread culling of small trees and understory shrubs to thin forests, the doctrine behind projects like Arizona’s ambitious 4 Forest Restoration Initiative, is a bad idea. This strategy, they say, will “move dry forests outside their historical range of variability, rather than restore them, probably with negative consequences for biological diversity."

Baker and Williams’ studies were met with strong opposition from some of the preeminent fire ecologists in the field. Thomas Swetnam, who directs the tree-ring research lab at the University of Arizona, said the team’s findings were inconsistent with a large body of published research, and called the papers “deeply flawed in multiple ways.” In an email he specifically critiqued the finding that high-severity fires aren’t any larger today than in the past.

“There is no reliable, convincing evidence that high severity crown fires creating canopy holes of this size occurred anytime in the past 500 to 1,000 years, probably much, much longer. These are extraordinary events, with major unsustainable, irreversible impacts,” he wrote, that can be explained by the conventional wisdom: unnaturally dense forests, drought and high temperatures.

Peter Brown, a fire ecologist and director of Rocky Mountain Tree-Ring Research in Ft. Collins, Colo., who studied under Swetnam, said he equated Baker with “creationists trying to tear down evolution." Ponderosa pine forest ecology, Brown said, is a theory, but one supported by studies from multiple disciplines. “You can’t just tear down that whole body of understanding and the complete set of evidence for that theory with one little bit of data that has its own tremendous flaws in it in the first place.”

The flaws, Brown says, stem from the Wyoming team’s use of GLO surveys, which he believes do not document a large enough area to provide an accurate depiction of the forest. But Baker says and his students walked hundreds of miles along the paths the surveyors traveled, looking for anything that would suggest their notes were wrong. He found that the trees the surveyors marked were 95 to 98 percent accurate.

Baker doesn’t disagree with Swetnam’s argument that current climate conditions are likely to make modern fires worse. He and Williams just hope to place the current rash of wildfire in its historic context. Big, severe wildfires are “a natural part of our forest ecosystems; we maybe didn’t know that until recently. But now that we do we just have to figure out how to live with that.”

Emily Guerin is an intern at High Country News.

Photos of the High Park Fire courtesy Flickr user U.S. Air Force.

Chas S Clifton
Chas S Clifton Subscriber
Jul 27, 2012 10:32 AM
Rather than surveyors' reports, I would put more trust in photographic studies such as <a href="[…]/B000PIF1EW">Yellow Ore, Yellow Hair, Yellow Pine</a>, which compared multiple photos of the Black Hills taken from the same spots in 1874 and 1974. You can really see how much thicker those ponderosa pine forests became over 100 years.
William L Baker
William L Baker Subscriber
Jul 27, 2012 11:47 AM
We appreciate HCN covering this topic, which obviously is controversial, but there is certainly a need for more research and more public discussion. There is much at stake, as these forests are among our most beautiful western forests, they are near many of our communities, fires are burning out of them into our communities, and there is no doubt that they are in need of ecological restoration after a century of intense logging and livestock grazing. Our research differs, however, in providing distinct messages that we hope will be carefully considered by scientists and the public. The evidence we have published shows that these forests historically burned at high-severity, meaning a large fraction of trees were killed. This occurred when fuels were at natural levels, before any of the widespread EuroAmerican land uses altered these forests. The implication is that today’s fuel-reduction programs that promise to restore fuels to their natural levels likely may not work, since severe fires were common when fuels were at these natural levels. Inside or near ponderosa pine forests or dry mixed-conifer forests thus are really bad places to live, even if fuel-reduction has been undertaken. Alternative approaches for western communities to protect themselves are in my book: Baker, W.L. 2009. Fire ecology in Rocky Mountain forests. Island Press, Washington, D.C.

Dr. Brown's comment about creationists is just sour grapes, probably because he does not like the findings of our research. He knows very well that our research is published in top, peer-reviewed scientific journals, the same kinds of scientific journals where he publishes. These scientific journals do not publish creationism. The journals are well-respected, including Ecological Monographs and Ecosphere, published by the Ecological Society of America. Our research was funded by the National Science Foundation and the U.S. Department of Agriculture, which have extremely competitive peer-reviewed grant programs. Dr. Brown’s comment is unfounded and also denigrating to the scientific enterprise in general and the many scientists who contribute to these programs and scientific journals.

Also, the dataset is not tiny, as Dr. Brown implied. There are >13,000 records from the surveyors across about 4 million acres of land in three states in the central study, which is published in Global Ecology and Biogeography (see below). Moreover, the records are as good or better than the kinds of data Dr. Brown and Swetnam rely upon. Even in the most heavily sampled area in Northern Arizona (Abella and Denton 2009), the tree-ring data that form the basis for our study (75 trees/ha) are more abundant than in tree-ring studies (69 trees/ha) and much more abundant than government weather data and government forest-monitoring (FIA) data that are considered a sound basis for scientific understanding of climate and forests. Also, the accuracy and validity of the survey data were validated in an extensive scientific trial, which was published in 2011 in Ecological Monographs, one of the top ecological science journals in the world. Moreover, the tree-ring data that Drs. Brown and Swetnam rely upon are very spatially limited, generally from small, isolated studies totally at most thousands of acres, not millions of acres as in our study.

Dr. Swetnam is incorrect. We published new evidence that the size of high-severity fire patches and the rate of high-severity fire is no different now than it was in a comparable period before EuroAmerican settlement, in natural forests. This study is in Ecosystems (see below).

Readers interested in seeing the actual evidence in these published studies can find them using the citations and links below, but may have to access them via subscriptions, or can email me at Bill Baker

Williams, M.A. and W.L. Baker. 2010. Bias and error in using survey records for
ponderosa pine landscape restoration. Journal of Biogeography 37:707-721.

Williams, M.A. and W.L. Baker. 2011. Testing the accuracy of new methods for reconstructing historical structure of forest landscapes using GLO survey data. Ecological Monographs 81:63-88.

Baker, W.L. 2012. Implications of spatially extensive historical data from surveys
for restoring dry forests of Oregon’s eastern Cascades. Ecosphere 3(3), article 23.

Williams, M.A. and W.L. Baker. 2012. Spatially extensive reconstructions show variable-severity fire and heterogeneous structure in historical western United States dry forests. Global Ecology and Biogeography, in press.

Williams, M.A. and W.L. Baker. 2012. Comparison of the Higher-Severity Fire Regime in Historical (A.D. 1800s) and Modern (A.D. 1984–2009) Montane Forests Across 624,156 ha of the Colorado Front Range. Ecosystems, in press.[…]amp;sortorder=desc&o=20
Steve Snyder
Steve Snyder Subscriber
Jul 27, 2012 12:45 PM
It seems to me that one reason some people might not like these new findings is they undercut the idea of just how well we can "manage" forests, especially at the urban-forest interface, where so many people with money want to move in the West.
Well, just maybe, we CAN'T manage that so well, and American exceptionalism takes another ding. It's called "acceptance," per the quote from Baker at the end of this story.

In that case, more states need to follow California's lead and institute fire protection surcharges for people living in those areas. Insurers may want to raise their rates. And, thus, maybe not so many people will move to these areas.
Wendy Beye
Wendy Beye
Jul 27, 2012 01:09 PM
And don't forget the 1910 "Big Burn" in northern Idaho and western Montana that burned 3 million acres of forest, killed 87 people, and destroyed numerous towns in a few days. That disaster prompted the fledgling Forest Service to begin trying to snuff out fires while they were small, using mules to carry supplies, hand tools, and firefighting crews deep into the forests.
Liana Aker
Liana Aker
Jul 27, 2012 03:50 PM
This is certainly compelling research and I would need to read the study to be fully informed but do the author's know the temporal regime of the historic fires and whether they were human-started fires or were all lightning-related?
Barry Pryor
Barry Pryor Subscriber
Jul 27, 2012 10:19 PM
I was told several years ago that sediment studies in Lake Tahoe confirmed the fact that big catastrophic fires were a natural part of the ecosystem. I don't know if this information was reliable or not, but it did make me wonder about the kinds of data that support one hypothesis or another.
Peter Brown
Peter Brown
Jul 29, 2012 06:55 AM
I would like to briefly follow up on my comments to Ms. Guerin with first an apology concerning my ill-chosen comparison of this recent work from Drs. Williams and Baker to creationists’ efforts: Certainly W&B are publishing their research in peer-reviewed journal articles and are obviously following scientific methods and ethics, in very strong contrast to the creationists. I sincerely apologize to both Drs. Williams and Baker for that comparison.

However, my point was that we have been here before. This new data set is, in my opinion, just one more in a series of approaches that appears from the outside to be part of an agenda to tear down any efforts - however well-meant, however well-documented - at ecological restoration of surface fire regimes in ponderosa pine ecosystems around the western US. So to my follow-up question about the data presented in their work: Even if their model of forest structure based on the GLO data is correct, there is still a grand leap from there to fire behavior. Current fires in ponderosa have contained large portions of active crown fire, with complete tree mortality over 1000s or even 10000s of acres. Tree recruitment into recent fires on the Front Range (e.g, Black Tiger, Buffalo Ck, Bobcat Gulch, Hayman, etc) as well as many other areas around the SW in particular has been either very slow or nonexistent. Considering the lack of seed sources that have occurred across some very extensive landscapes and the relatively large size and lack of mobility of ponderosa seed, how does the lack of recruitment after recent fires fit with their model of past fire behavior, if the current fires are within the HRV according to this model? Most certainly current forests are not coming back as dense stands of even-aged trees. Right there something is incomplete in the model, and strongly suggests that their interpretations of the GLO data - and hence implications contained in these data for ecological restoration efforts - are flawed.
Tim Baker
Tim Baker Subscriber
Jul 29, 2012 09:37 AM
It is in the best practice of science that widely accepted paradigms are challenged from time to time if for no other reason that it forces us to step back and really examine what we think we know. I think the research mentioned here certainly falls into that category as we work to expand our understanding of fire processes in western ecosystems.

The biggest difficulty I'm having with the conclusions of Drs. Baker and Williams is that it seems counter-intuitive to the evolutionary selection of traits in ponderosa pine -- thick bark, fast juvenile growth, and good self-pruning -- which seem tailor made for a low-intensity fire regime. As opposed to lodgepole pine -- thin bark, early maturation, serotinous cones, poor stem exclusion -- which would benefit from an infrequent high-intesity regime. And with mixed-conifer stands the story is more complicated of course, hence mixed-severity regimes being thought of as common in them (with the true firs in areas mostly devoid of fire).

It's this incongruity that I'm trying to reconcile with the idea that infrequent high-severity fires were the historic norm across a wide array of western forests.
Bill Gore
Bill Gore Subscriber
Jul 29, 2012 11:06 PM
The Rule of Thumb, or the key to the forest 'fire mosaic' is that only the strong survive. Especially in Ponderosa stands that characterize the Mogollon Rim and the eastern Cascades, immature trees are consumed by periodic fires, while mature giants shrug it off..Our attitude towards fire is still shaped too much by european cultural attitudes about the landscape and our place in it. In areas of expanding urban-wildland interface, such as the front range (recently in the news) our affluent, dispersed suburban mode of settlement, and our demands for total safety of life and property, have seriously interrupted the normal fire regime. The BEST fire regime for the landscape is the one that happens...What happens, happens..
The longer the actual event is postponed, the more catastrophic and dangerous the actual event. Natural fire regimes are actually pretty low key. No big deal, just a part of summer (or Fall in California).
Clark Stoner
Clark Stoner
Jul 30, 2012 12:50 PM
Although clearly a clever use of GLO notes, I think the GLO survey notes should be read from the point of view of the surveyors who wrote them and in the context of the times.

The GLO surveyors weren't exactly welcomed guests in those days. The U.S. Government was at war with various indian tribes at various times across the west. There is little doubt that the indians understood the surveyors preceded the western movement of the white man, which they likely considered threatening. The survey parties weren't exactly all that well paid, and they likely often feared for their safety. So, I would have taken a rather tongue in cheek approach to the reliance of the GLO notes when they were describing the landscape (they most often served future boundary retracement surveyors needs quite well). In my experience with reading GLO notes in California and Nevada, dating back from the 1860s through the 1880s, I find that they weren't exactly describing forested land in terms of any measurable density. Their descriptions of forested land were actually quite general and subjective. You were taking a gamble if you were a speculator in the GLO office trying to decide whether or not you wanted to buy the land described. So in the study it appears they found quite a number of bearing trees (which is cool), but their existence is no testament to the density of the forest at the time the GLO surveyors marked them.

Also, why is reliance on data from the 1880s so relevant? I seem to recall Lewis and Clark witnessing the indians setting fire to the landscape in 1804, which I believe it is now accepted that the indians were performing their own forestry management practices prior to the arrival of the Europeans, and most agree their practices were to promote fertile hunting grounds. So, as the indians got wrapped up in wars with the U.S. Government, it might be safe to say that their normal forestry management practices were disrupted and in various cases the forests appeared quite different by the time the GLO surveyors arrived. The GLO surveyors in many cases might not have been witnessing the forests in their centuries prior "pre-existing" state (saving the term "natural state").

Interesting article, interesting study. I am left with only questions.

Clark Stoner, PE, PLS
William L Baker
William L Baker Subscriber
Aug 14, 2012 09:23 AM
Thanks to readers for interest in our study. Please continue to ask questions or provide comments on it. Here are a few responses:

1. Tree recruitment after high-severity fires in dry forests appears to have at times been slow historically as well as recently. Yes, our methods may underestimate the amount of high-severity fire in unusual cases where trees regenerate very slowly and sparsely after fire, as we mention in the GEB article in the paragraph right before the Conclusion. However, this means that there likely was somewhat more high-severity fire in historical dry forests.

2. The agenda to stop restoration of low-severity fire is just as imaginary as was the creationism agenda. See the Ecoscience article’s last paragraph, for example.

3. The traits of ponderosa pine that confer resistance to low-intensity fires (e.g., thick bark, elevated crowns) would increase survival after these fires. However, ponderosa pine also has reasonable seed dispersal, is relatively drought tolerant, and grows well in the open, high-light environments found after high-severity fires, and thus is sufficiently adapted to these fires as well. This is evidenced by widespread regeneration of these forests after the high-severity fires we have documented from the 19th century.

4. Yes, surveyors operated under difficult conditions. However, the surveyor records are remarkably accurate, as documented in our Journal of Biogeography article. We were able to use the survey notes to relocate many surviving trees, a testament itself to surveyor accuracy, and remeasure them with modern accurate instruments. Forest density was not reconstructed from general descriptions, but instead from these data for the bearing trees marked and measured by the surveyors at section corners. The Ecological Monographs study explains in detail how all of this was done.

5. Forests at the time of the surveys may have been affected by burning and other activities of Indians, but the impacts of burning by Indians have often been overstated and simplified. For a comprehensive review of Indian impacts, see: Vale, Thomas R. (Editor) 2002. Fire, native peoples, and the natural landscape. Island Press, Washington, D.C. 1. Fire causes are not known, but forests at the time of the surveys were little affected by fire exclusion, logging, and other widespread EuroAmerican land uses.

Bill Baker, Univ. of Wyoming