- Author: Rob York
Article Reviewed: Do mountain pine beetle outbreaks change the probability of active crown fire in lodgepole pine forests?
This review provided by the Battles lab of UC Berkeley
By M. Simard, W.H. Romme, J.M. Griffin, and M.G. Turner. Online preprint. Ecological Monographs. Availability: http://esa.org/papers/
Plot line: Bark beetle outbreaks have caused extensive mortality of pine forests across western North America. In the aftermath, these forests with many dead and dying pines are widely considered to be at extreme risk of catastrophic wildfire. The goal of this study was to evaluate this conventional wisdom. Specifically they focused on the potential for the combined effects of these two disturbances, beetle kill and fire¸ to alter the structure and function of the Greater Yellowstone Ecosystem. Their approach was to quantify the impact of beetle-induced mortality on fire behavior in lodgepole pine forests. They measured forests fuels in an extensive network of plots that spanned a gradient in time since bark-beetle mortality. In other words, they used a chronosequence approach. This chronosequence included currently undisturbed stands (no bark-beetle mortality) to stands where bark beetles killed all the pines in the canopy 36 years ago. They then used this empirical data to inform fire behavior models. These models were used to simulate the impact of beetle damage on fire behavior.
The results of the simulations convincingly showed that mountain pine beetle outbreaks in the lodgepole pine forests of Greater Yellowstone do not increase the risk of active crown fires, the most destructive type of wildfire. Instead, their results suggest that beetle kill may actually decrease the risk of crown fire. Their explanation is that the main fuel for crown fires is tree needles. Once killed by mountain pine beetles, these needles are retained in the canopy for only brief period (1-2 years post outbreak). Thus the exposure to the risk of catastrophic crown fires is short. In the longer term (25-35 years post outbreak), the accelerated growth of understory trees may increase the potential for passive crown fires. However these fires are less intense and spread more slowly than active crown fires. In general, the authors concluded that weather conditions may have a greater influence on wildfire behavior than fuel characteristics affected by bark-beetle mortality.
Relevant Quote: “Our results suggest that mountain pine beetle outbreaks in Greater Yellowstone may reduce the probability of active crown fire in the short term by thinning lodgepole pine canopies.”
Relevance to landowners/stakeholders
The public perception of the ongoing bark-beetle outbreak in western North America is that the resulting forest of dead red and grey trees is an environmental disaster. Chief among the concerns is the risk of catastrophic wildfires like those experienced in 2007 in southern California (primarily San Diego county). This study, one of the few informed by evidence, suggests that the perceived risk of wildfire is exaggerated for the lodgepole pine forests in the Yellowstone area. However this study does not address other forest health concerns related to the widespread tree mortality nor does it address the future dynamics of these forests.
Relevance to managers
The abundance of dead and dying in trees in the western North America causes great concern among private and public forest managers. These managers struggle with the question of what to do about the “aftermath” forests. One option is salvage harvesting. The arguments for and against such an intervention are complex but a common reason to intervene is to reduce the risk of catastrophic wildfires. The results from this research cast doubt on this rationale. Based on their data and simulations, beetle kill does not increase the risk or hazard of wildfire in these forests.
Critique and/or limitations (there’s always something, no matter how good the article is) for the pedants:
This research relies on two well-established methodologies in forest science – chronosequences to acquire temporal data and simulation models to predict ecosystem behavior. Chronosequences assume that the only difference between research sites is time since last disturbance (in this case, the disturbance is the bark beetle outbreak). The authors do a good job supporting this assumption but still no space-for-time substitution is perfect. Fire science must rely on models since direct tests at the appropriate scale are impossible. Nobody is going to set an experimental fire at a landscape level. The fire behavior simulator used in this paper is a well-established approach. The physics of fire are well understood. However our ability to accurately represent the three dimensional distribution of fuels in a forest is limited. For example, the surface fuel loads and their interaction with fire are summarized in fuel models. While the authors did measure the surface fuel loads, they did not explain how they translated these results into synoptic fuel models. Nor do they specify which fuel models were used in their simulations. Also a key to their argument is the potential for crown fires in the aftermath forests but the current fire models (like the NEXUS model used in the paper) struggle to capture the complex dynamics of crown fires. This challenge is particularly relevant to question posed in the title since mountain pine beetle outbreaks kill trees while they are still standing. There is a well-documented progression from recently dead trees where all the leaves die and turn red but remain on the tree (red stage) to the later stage where these dead leaves have fallen off the tree. Leafless standing dead trees are referred to as the grey stage. The transition from live green tree to red stage to grey stage also represents a change in canopy fuel loads. The authors account for these losses in foliage by assigning discount rates but these assignments are made in an ad hoc manner without support or evidence to justify the discounts. Since predicting crown fire behavior is essential to assessing the hazard related to beetle-induced mortality, it seems that a sensitivity analysis is needed to demonstrate that the results are not particularly dependent on the discount rate assignments.