More evidence of the “fireproofing effect” of insect outbreaks in a forest

Additional research finds evidence of a “fireproofing” effect on host trees from defoliation due to western spruce budworm outbreaks.

western spruce budworms beetles

Above: Grand fir in Oregon defoliated by western spruce budworms. William M. Ciesla.

There is no dispute that severe outbreaks of western spruce budworm (WSB) and mountain pine beetle (MPB) in a forest have huge visual impacts. Many land managers have worried about more, larger wildfires and politicians have used it as an excuse for more logging.

But the commonly held belief that the effects will lead to higher intensity, more rapidly spreading wildfires has been disproven many times in the last eight years by scientists.

We first wrote about this issue in 2010 (Firefighters should calm down about beetle-killed forests) when some early research started to bring the facts to light.

The WSB and MPB attack trees very differently. The WSB defoliates the tree, consuming the needles and removing fuel from the canopy relatively quickly. The MPB kills the tree from the inside, leaving the dying “red” needles on the tree until they fall off in one to two years. The possibility of crown fires may increase during that red needle period, but it makes sense that fewer fine fuels in the canopy would reduce the fire intensity and make it less prone to transition from a ground fire to a crown fire. Both types of attacks eventually produce more course fuel on the forest floor as the branches break off and  the trees eventually fall over.

Research conducted by Daniel G. Gavin, Aquila Flower, Greg M. Cohn, Russell A. Parsons, and Emily K. Heyerdahl found evidence of a “fire proofing effect” for outbreaks of WSB. It does not address the MPB.

Below is an excerpt from their work:


“Extrapolating Results: Reduced Tree Mortality

“Taken together, the tree-ring and modeling studies suggest a lack of synergism between WSB outbreaks and wildland fires. However, a different kind of synergism may exist: Defoliation might dampen the severity of a subsequent wildfire. To explore this possibility, we used existing empirical equations that show the probability of mortality due to defoliation (fig. 3A) and the probability of mortality due to crown scorch (fig. 3B), combined with the simulated results of canopy consumption at different levels of defoliation (fig. 3C), to extrapolate the summed probability of mortality under a range of surface fire intensities and defoliation levels (fig. 3D). The results suggested a distinct “fireproofing” effect of defoliation: The increased risk of mortality by WSB is more than compensated for by reduced foliage consumption during moderate surface fire intensities. For example, trees with 50-percent defoliation have a distinctly lower probability of mortality when surface fires are less than about 74 kilowatts per square foot (800 kW/m2 ).

“However, we considered only the partial effect of defoliation on fire occurrence; we did not take into account other effects of WSB outbreaks, such as mortality of small trees. Of course, field observations are required to test our prediction. Remotely sensed burn severity maps, in combination with prior surveys of insect effects, could address this issue. One such study of the 2003 B&B Complex Fire in Oregon showed that prior defoliation had a marginal effect on reducing fire severity that was not statistically significant (Crickmore 2011). However, an analysis by Meigs and others (2016) of all post-WSB fires in Washington and Oregon from 1987 to 2011 showed that there is a statistically significant reduction in fire severity that persists for up to 20 years following an outbreak. Thus, the effect of defoliation on crown fire behavior modeled by Cohn and others (2014) appears to be confirmed by the analysis of burn severity data by Meigs and others (2016).

“Fireproofing Effect?

“It may seem reasonable to assume that extensive defoliation, causing sustained low levels of tree mortality in mature trees, should have a measurable effect on wildfire occurrence. However, fire is a highly variable disturbance in itself, and it is highly sensitive to specific climate and winds during the fire event. The scale of fuel changes wrought by WSB may be too small to affect subsequent fire probability in ecosystems where fire is limited by fuel moisture and ignition sources rather than fuel availability. Our data show that these two disturbance types do not share similar histories, despite a common link to drought events.

“Nevertheless, we hypothesize a “fireproofing” effect on host trees from defoliation due to WSB outbreaks. Although such an effect has been detected statistically from recent fire events (Preisler and others 2010; Meigs and others 2016), the inferred processes at play remain to be studied in detail at the site scale.”


UPDATE January 5, 2018: These researchers are not the only ones with similar findings on this subject. If you would like to read more, scroll through the articles on Wildfire Today tagged “beetles”.

Author: Bill Gabbert

After working full time in wildland fire for 33 years, he continues to learn, and strives to be a Student of Fire. Google+

9 thoughts on “More evidence of the “fireproofing effect” of insect outbreaks in a forest”

  1. Oh dear! With all due respect, they didn’t do any actual burns to test their unproven hypothesis. It’s all computer modeling, with estimates and probability suppositions based on guesstimate formulas, with absolutely no real life trials of any kind.
    The forest conditions are very serious, a false sense of security based on an incomplete and wholly theoretical computer model could be disastrous. Safer to presume any fire in those zones will be terrible rather than banking on maybe it’ll be OK? Also, just curious, anyone know what’s that sort of weird moss hanging off the dying trees?

  2. Has anyone studied the impact of hemlock woolly adelgid and balsam woolly adelgid damage on fire conditions in the east, especially in the Smokies? It seems wildfires in the east are no longer a non-issue.

  3. Color me unsurprised. And probably anyone else who has spent enough time on fires to understand that fine fuels are an essential driver in determining fire spread. Fewer fine fuels (like after dead conifers drop their needles) will lead to slower fire spread rates which will lead to decreased resistance to control.

  4. Statistical lab based mathematical analysis of past bug out breaks which has not been ground trushed is vodo based science. Spend 35+ years actually fighting both low and high intensity/ severity wildfires in the western USA before believing that the lack dead wildland timber fuel types of any size class during sever drought and fire behavior/fire danger ratings doesn’t significantly contribute to wildfire intensity, resistance to control, and severe fuels driven Fire behavior. More fodder for extremist who don’t believe in long term groundbtruthed real science based forest management like thinning and removing dead and dying timber. Saw that theory go out the window once again this summer when I was asked to come out of my 3 year old retirement as an OPSC1 on the Lolo Peak and Rice Ridge Fires in W MT. But what do I know? I got a “D” in statistics in college in 1982. Maybe for a good reason!

    1. Sir, they don’t have actual ‘findings’ , they have a hypothesis. The scientific method starts with a ‘Hypothesis,’ which is then tested to determine if there is truth to it which then allows it to be categorized as ‘Theory’ then after further testing may fulfill the scientific requirements of ‘Law.’ Most of our science that is ‘Law’ has been proved in physics, so much of anything non-physics that we rely on is actually Theory – it’s got a demonstrated truth that’s reliable enough to base many plans and actions on for success.
      Most hypothesis end up being disproved in the process of testing, some are proved to be classed as Theory. In this case, everything they have presented is only un-tested hypothesis, based on a review of other people’s published work and some hypothetical computer modeling, which is based on equations they created. It seems to have more authority (i.e. real conditions scientific testing data which have to be from actual real burns to see if the conditions they are predicting actually do happen) than it has.

      Not making a determination as to whether their hypothesis is correct or not. We don’t know. Critically, a careful reading of their words indicates that they admit that they don’t know either. Their predictions might be correct, but might not be. They might be proved correct part of the time under certain conditions, but not in others. After actually standing in one of these afflicted areas, we doubt the stated hypothesis, but in the same way, we haven’t tested our hypothesis either so we don’t know if our on-site ground assessment is true or partially true or not true.
      They could be right; fire conditions sometimes seem counter-intuitive, especially to the non-professionals, but given the vast expanses of so much dead debris and timber, caution seems prudent. Sadly, perhaps some proscribed burns years ago on the perimeters could have stopped the infestations from spreading over such vast tracts. On a positive note, most of the areas are too rugged and steep for any human populations/habitations to worry about.
      Unfortunately, an actual fire will probably give firefighters more real-life information than anyone ever wanted.

      1. Anyway, do appreciate very much reading more about this topic, as the immensity of the damage in Colorado is a gutting sight to the newcomer. (Obviously many people have been studying this for years, so still learning about it).

  5. It’s worth noting that the eastern spruce budworm situation, at least in Ontario, is a considerably different situation. See for example the article published in the Forestry Chronicle in 1987 by Brian Stocks of the Canadian Forest Service which presents empirical data on the fire behaviour potential of eastern spruce budworm-killed balsam fir stands as derived from outdoor experimental burning in north-central Ontario: http://cfs.nrcan.gc.ca/publications?id=22429

    One needs to be cautious of course about generalizing.

    Marty Alexander

Leave a Reply

Your email address will not be published. Required fields are marked *