Rethinking Forest Health

George Wuerthner

I just read through a portion of the Beaverhead Deerlodge National Forest (BDNF) revised plan. Among the major components of the plan is support for “vegetation management,” a euphemism for logging. The BDNF plan calls for “treating” its forests by logging to “restore” its ecological health. It has become commonplace for the Forest Service to justify logging for forest health reasons instead of timber production. We no longer log just to get the raw material for lumber and profits for timber companies. We log the forest to restore ecological health, or so the agency suggests.

I personally don’t believe that the BDNF staff is purposefully using “forest health” as an excuse to log. There is a wide-spread assumption among many forest ecologists that past forest management, including past logging, along with fire suppression, has radically altered our forests. However, the agency may be unaware of more recent research that calls into question many of these previous assumptions about forest condition and health.

Even if the assumptions about forest condition are correct, that doesn’t mean that logging can actually restore the presumed “historic range of variability.” One could restore ecological health by permitting more fires to burn, and by the use of more prescribed burning. Since this doesn’t produce profits for the timber industry, the agency is under a lot of pressure to cut trees instead of using less intrusive means like prescribed burning and wildfire as a means of restoring the presumed forest conditions. To its credit, in its Alternative 3 of the forest plan the BDNF does recommend exactly that prescription—more wildfire and prescribed burning and limited logging. Unfortunately, for the public, Alternative 3 is not selected by the agency as its preferred alternative.

The problem for anyone advocating “restoration” is that we have few references about how the forest looked a hundred years ago. There are some historic photographs that provide a valuable perspective, but whether these represent just a point in time and at a particular spot, or are characteristic of the forest as a whole is unknown. Furthermore, there is always the potential for a selective bias in the choice of photographs by the researcher seeking to find evidence for a change in forest condition and composition.

The same can be said about written accounts. When someone asserts that the forests were so open they could ride a horse through them could again reflect a bias in the observer who either selected the easiest pathway through the woods, avoiding other denser forest stands, or even a failure to note when the forests encountered were densely forested. Also there is always the chance for researcher bias that ignores some references to forest condition, in favor of descriptions that fit one’s preconceived notions about how the forest appeared.

The further back in time you go, the murkier the record. Most ecologists must rely upon reconstruction of past “historic conditions” by proxy. One popular method involves looking at fire scars on trees, and trying to determine past fire intervals. The assumption is that low intensity fires do not kill trees, but rather leave a record of their occurrence by a scar. By reading the intervals between such fire scars, researchers can reconstruct past fire occurrence and severity and make some assumptions about the historic look of these forests. However, a recent review of this method by a number of researchers has called into question the validity of many of these studies.

For instance, William Baker from the University of Wyoming and colleagues did a review of fire history studies in ponderosa pine forests and found that nearly half of them depended upon only one or two trees. Such a small sample size is suspect. Furthermore, even when a larger sample is used, there is a tendency for fire researchers to sample trees where there is an abundance of fire-scarred trees. However, such a bias in sampling may not represent the historic conditions of the forested landscape as a whole. Baker’s research suggests that the occurrence of stand replacement fires may have been greater than previously assumed, even for low elevation dry forests.

Another study done by Forest Service researcher Paul Hessburg and associates looked at the temporal patterns of eastside forests in the Cascades. He started with the assumption that past conditions would be reflected by the stand composition of the present forest. Using randomly selected air photos to review forest stand composition, he determined that there was little evidence for so called “light, low intensity” burns or “open park-like” forests in dry low elevation and moist mixed forests as presumed. Rather partial and stand replacement fires appeared to be the norm—even before fire suppression was effective and presumably created a “fuels build up.”

A third study in Colorado done by Dominick Kulakowski and his associates critiqued the Forest Service’s assumption that there was wide-spread “decline” in aspen. Kulakowski was fortunate in finding a highly detailed and accurate 1898 map of forest type and occurrence of recent burns for a portion of the Grand Mesa area of Colorado. Digitizing the map, and then comparing it to the present vegetation type for the forest, he was able to determine that relative to the late 1800s, a larger portion of the landscape was covered with aspen today than a century ago. A rash of fires near the turn of the century as a result of more favorable climatic conditions for fires (i.e. drought), as well as burning by sheep herders, miners, and other settlers contributed to an increase in aspen throughout the 20th Century. So measured against people’s recollection of aspen abundance in the recent past century, there had been a decline in aspen. But what Kulakowski’s research showed is that the current abundance of aspen was not outside of the historic range of variability—and conifer cover was actually greater a hundred years ago than today.

A fourth study of wildfires in the northern Rockies by Penny Morgan, of the University of Idaho, found one more piece of evidence that can be used to question the assumptions about “historic range of variability.” She mapped known wildfires on national forests in Idaho and western Montana from 1900 through 2003. She found the majority of all large fires occurred in just 11 fire years. These fire years coincided with extensive drought. The first six big fire years occurred prior the mid-1930s and the last five years have been since 1988—the year that much of the Yellowstone ecosystem burned. Between the 1940s through the late 1980s, moister conditions resulted in virtually no large fires in the entire region. This has major implications for our assumptions about fire suppression and fuels.

Many people use the recent past as their point of reference. In other words, people talk about the large fires we are experiencing today as compared to the 1940s, 50s, 60s and 70s and presumed that the reason has to be a consequence of greater fuels. But what is intriguing about her research is that six of the large fires occurred long before anyone can claim that fire suppression was responsible for a “fuels buildup.” No one can reasonably assert that fire suppression and fuel buildup was responsible for the huge 1910 Burn that raged across more than 3 million acres of northern Idaho and western Montana. Drought and wind drove those fires, as it has all recent big fires.

The more recent spate of large fires in the 1990s and 2000s are attributed to “fuel buildup” as a consequence of this fire suppression. However, the recent period of large fires also coincides with historically severe drought conditions across the West—the kind of climatic conditions that has always driven large blazes. Severe drought and overall warmer temperatures are also responsible for widespread beetle outbreaks. Beetle experts, however, do not see the large die-off of trees due to beetles as out of the ordinary—and many assume that such large scale beetles outbreaks have occurred in the past, again calling into question the assumption that our forests are “unhealthy.”

Temporal scale is an important factor in how we view current conditions—the longer the time frame of reference, the less current conditions seem unnatural. A study by Boise State University professor Jen Perce and colleagues looked at fire frequency and scale among ponderosa pine forests along the Payette River in Idaho. Using the geological fire history recorded by charcoal buried in soil sediment, she concluded, contrary to popular perception that low intensity blazes are the norm for low elevation dry forests, when viewed over longer time scales, climatic conditions like drought has led to significant stand replacement fires on occasion, even in ponderosa pine ecosystems.

What do all these studies and others suggest about the presumed “historic range of variability”? The message I take from these studies is that climate controls big fires and, when viewed on a landscape scale, our forests may not be out of balance as presumed. In fact our forests are very healthy and what we are seeing with both large blazes and large scale beetle outbreaks are within the “norm” for these forests if climatic conditions are taken into account. The large fires we are experiencing are “resetting” the ecological parameters of the region. There is no need to “restore” forest health—the forests are perfectly healthy and are restoring themselves—without the help of the timber industry, thank you.

Furthermore, even if it can be proved that some forests are somewhat out of “balance” that doesn’t necessarily mean that intrusive logging is necessary or can restore forest health, especially since logging has many other negative impacts that are often ignored or glossed over. These include the creation of access roads that decrease habitat security for wildlife, act as vectors to spread weeds, not to mention are a major source of sedimentation into streams (sedimentation from fires is short lived—while roads “leak” sediment for decades).

Logging operations seldom leave as many snags as naturally occur as a result of fire or beetles. Logging also removes snags which are critical to the survival of many species—for instance; more than a third of all birds in the northern Rockies are cavity nesters, not to mention use of snags by a host of other species from bats to snails. Plus, logs charred by fires take longer to decompose and last longer as a structural component in the ecosystem—with long term consequences for wildlife and nutrient flows. The presumption that logging “emulates” nature is a bunch of timber industry propaganda.

Finally, new research is calling into question the other major justifications for logging which includes the assertion that logging can stop or reduce large fire risk and/or insect outbreaks. Logging does not affect the conditions that drives large blazes namely drought, high temperatures, low humidity, and, most importantly, wind. In fact, there is even evidence to suggest that thinning the forest can substantially exacerbate these conditions leading to increased solar drying of fuels, and permitting greater penetration of wind. Even a five mile an hour increase in wind results in an exponential increase in fire spread. And removal of competing trees, leads to rapid regrowth of shrubs and smaller trees that are more flammable. The best way to reduce fire risk to communities is to fire-proof homes, not the forest.

Circling back to the BDNF plan, all of this research calls into question the Forest Service assumptions about what is “normal” for the BDNF as well as many other forests in the region. It is possible that the Forest Service assumptions about the forest conditions are accurate. On the other hand, there is more than a reasonable likelihood that our forests are well within the “historic range of variability” and need no intrusive management other than to get out of the way and allow fires, beetles, droughts, and other normal ecological processes to operate.