Comment on Hanson, C.T. Cumulative Severity of Thinned and Unthinned Forests in a Large California Wildfire. Land 2022, 11, 373

Abstract

Hanson (2022) evaluated thinning as a management tool to reduce wildfire tree mortality using thinning history and fire mortality data from the Goosenest Adaptive Management Area (GAMA) in California. The combination of on-the-ground measurements of thinning tree removal with remote sensing fire mortality data is questionable. Using his data, fire mortality in thinned stands was calculated and found to be less than mortality in unthinned (control) stands. Fire mortality in experimental plots does not properly represent the full benefits of density reduction because lower density provides a risk reduction that increases as more area is treated, due to the nature of fire spread. Thus, the conclusion that thinning is not beneficial is not supported by Hanson’s study and does not consider the importance of individual tree mortality vs. catastrophic forest loss and associated reductions in ecosystem services and loss of human infrastructure.

1. Introduction

Hanson [1] evaluated thinning as a management tool to reduce wildfire tree mortality. The Goosenest Adaptive Management Area (GAMA) in California had plots established in 1998–2000 with commercial thinning and controlled burning with paired unthinned stands (further discussed in [2,3]). The GAMA subsequently experienced a wildfire in 2021, which allowed for a unique opportunity to evaluate the effects of thinning on wildfire tree mortality. Hanson compared estimates of basal area before and after thinning obtained from [2] with estimates of basal area after the 2021 wildfire. He counted basal area removed by thinning as part of total mortality. He concluded that more “total tree death” following fire resulted from thinning than in unthinned stands (i.e., control plots). He used this result to argue that thinning is not effective at reducing fire risk. Here I argue that isolated stands are at an incorrect spatial scale for drawing this conclusion, that the data suggest that thinning actually provided protection for the remaining trees following thinning, and that an error in his table slightly alters his conclusions.

There are some important methodological aspects that may have also affected results. The removal of trees by thinning was estimated with data from a prior study using on-the-ground accounting of harvests [2], while fire mortality was based on Rapid Assessment of Vegetation Condition after Wildfire (RAVG) satellite imagery. These two techniques likely vary in data type (e.g., basal area as a continuous variable versus the basal area class system used by RAVG), and in both precision and accuracy. It is neither obvious nor demonstrated in this study that these two different methods can be combined in an analysis, particularly without taking into consideration the error surrounding each of these data collection techniques. In addition, fire mortality was based on data obtained a few months after the 2021 wildfire, whereas trees may continue to die for some years after a fire [4]. Because thinning in the GAMA shifted forest composition toward larger trees and pines versus the more fire-sensitive white fir [Abies concolor (Gordon) Lindl. ex Hildebr], more trees would be expected to show delayed mortality in the control (unthinned) plots.

2. Analysis

Thinning is a common silvicultural practice in commercial stands. It is often used to increase growth of remaining trees due to reduced competition, which increases their future financial value, and to remove damaged and diseased trees. An additional benefit might be reduced fire risk, which can be defined in several ways. In the case of a commercial thinning (which means the thinning at least paid for itself via the value of harvested trees), the risk consists of damage to the post-thinning stand. Using Hanson’s Table 1 [1] and taking his methods at face value, I backed out the post-thinning fire losses. Thinned site 19-3 shows no thinning basal area loss (value is 0 in column 3). Because this appears to mean that either this plot was not actually thinned or that there was a data-entry error, this plot was dropped. The mean fire-caused basal area loss for the remaining 26 plots was 22.5% vs. 39.6% for the thinned and unthinned plots, respectively. Thus, unthinned plots had 17.1% more basal area loss, an increase of 76%, but the difference was not statistically significant by a t-test because the sample size was not large relative to the variability among stands. Given that 21–23 years had passed since thinning, allowing understory regrowth, as shown by Ritchie [3] even by year 14, the potential protection for more recent thinning operations would likely be higher. Viewed in terms of protecting a thinned stand, there was thus a benefit from thinning. Because the more valuable trees are typically left and residual trees have had (and will have) increased growth rates, the financial benefit is greater than the 17% estimated basal area mortality benefit of thinning.

3. Discussion

Given the issues with small sample size and potentially incompatible methods, it is difficult to conclude that thinning does or does not reduce tree mortality. Tree mortality is also not the only concern in a commercial thinning operation.

Experimental thinning of stands, as in the GAMA, does not tell the full story on benefits. Fire spreads from spot to spot, so the condition of the whole forest matters for both fire size and intensity [5]. If a fire is already in the crown when it approaches a thinned stand, it may well stay in the crown and kill lots of trees. If a fire is burning in the understory in one stand it may stay low as it enters the next stand. Thus, a fire in a forest with many thinned stands (or low density) will more likely not become a crown fire, under similar environmental conditions, due to greater spacing among trees. This is evident in the fire history of ponderosa pine forests in much of the US West. Before the mid-1800s, frequent low-intensity fires burned through these forests without destroying them [2,6]. These fires were likely set regularly by Native Americans pre-colonial European colonization. Low-intensity fires do not create firestorms which spread burning debris far and wide, thus making the fire larger. They are easier to control. Firebreaks will work in a low-intensity fire, whereas in a crown fire they do not. Extensively thinned or low-density forest will thus be less likely to experience a mega fire and will be less dangerous to human structures and lives. This adds to both the financial and ecological benefits of thinning or density reduction compared to the benefit from treating single isolated stands.

Finally, it is useful to ask in what sense it is helpful to combine thinned tree removal with fire mortality as a metric of “total” mortality. This seems to presume that a count of trees per se is an ecological benefit, but Hanson does not document any such benefit. Thinning and logging, in general, are a major commercial business. Thinning provides forest health benefits (via removal of diseased trees), revenue from the thinning, and increased future revenue by allowing the remaining trees to grow faster, as Ritchie [3] demonstrated in this case. Simply allowing forests to burn is not good forestry and harms ecosystem services and human environments.