Fuel treatment effects on tree-based forest carbon storage and emissions under modeled wildfire scenarios
Matthew Hurteau and Malcolm North. 2009. Fuel treatment effects on tree-based forest carbon storage and emissions under modeled wildfire scenarios. Front Ecol Environ 2009; 7, doi:10.1890/080049
Full text [here]
Selected excerpts:
Abstract
Forests are viewed as a potential sink for carbon (C) that might otherwise contribute to climate change. It is unclear, however, how to manage forests with frequent fire regimes to maximize C storage while reducing C emissions from prescribed burns or wildfire. We modeled the effects of eight different fuel treatments on tree-based C storage and release over a century, with and without wildfire. Model runs show that, after a century of growth without wildfire, the control stored the most C. However, when wildfire was included in the model, the control had the largest total C emission and largest reduction in live-tree-based C stocks. In model runs including wildfire, the final amount of tree-based C sequestered was most affected by the stand structure initially produced by the different fuel treatments. In wildfire-prone forests, tree-based C stocks were best protected by fuel treatments that produced a low-density stand structure dominated by large, fire-resistant pines. …
As trees grow, C is sequestered, and these additional tons of C can be used to offset emissions in other sectors. In fire-prone forests, however, tree-based C storage may lead to large releases of C if trees are killed and partially consumed by a high-severity fire (Breshears and Allen 2002; Hurtt et al. 2002; Kashian et al. 2006; Hurteau et al. 2008). …
Beginning in the mid-1900s, US forested lands became a net sink for CO2, as a result of forest regrowth and fire suppression (Hurtt et al. 2002). Fire suppression has increased forest density and stand-replacement fire risk in forests that were historically characterized by frequent, low-severity fire regimes (McKelvey and Busse 1996). …
In fire-prone forests of the western US, there are three common management practices for reducing forest biomass and the risk of catastrophic fire: prescribed fire, mechanical thinning, and both treatments combined. …
Our objective was to model the amount of live- and dead-tree-based C stored and released over a century with and without wildfire in Sierra Nevada mixed-conifer forests, after fuel reduction treatments. Our hypotheses were:
(1) in the absence of wildfire, the no-fuels treatment alternative will store the most live- and dead-treebased C;
(2) with wildfire, treatments that develop and retain large trees will store the greatest amount of livetree C;
(3) pre-settlement forest structure will maximize tree-based C storage while minimizing C release during wildfire;
(4) with wildfire, prescribed fire treatments will have a lower total C release than unburned treatments; and
(5) reducing stand density and concentrating live tree C stocks in larger individuals will decrease the post-wildfire mortality, reducing the drop below the baseline.
Here, we use current CCAR FSP (2007) accounting methods to evaluate changes in C stocks using the Forest Vegetation Simulator (FVS) and track fire emissions using the Fire and Fuels Extension (FFE) of FVS (Crookston and Dixon 2005). Although FVS does not account for soil C, it is regionally calibrated, widely used by managers to model forest response to different treatments and disturbances, and one of the CCAR-approved models for establishing baselines. …
Methods
We used data collected in the Teakettle Experiment (http://teakettle.ucdavis.edu), in which all trees > 5-cm diameter at breast height (dbh) were measured and mapped in 18 replicate 4-ha plots. Using FVS, we modeled the effects of eight treatments (control, burn only, understory thin, understory thin and burn, restoration thin [based on North et al. 2007], restoration thin and burn, 1865 reconstruction, and 1865 reconstruction and burn) on tree-based C stocks. …
Results
… High-density stands, such as the control, had a higher proportion of their total aboveground C in dead biomass than did open stands (ie the 1865 treatments), which had a greater proportion of large-diameter, fire-resistant species.
Wildfire emissions were highest in the control (Figure 2) and decreased in order of understory, restoration, and 1865 treatments. Thinning treatments that included prescribed fire had lower wildfire emissions than did treatments that only involved thinning (Figure 2). Stands with a higher percentage of dead biomass had higher wildfire emissions (Figure 1). Over five applications, total prescribed burn emissions were 2–3 times higher than one-time wildfire emissions in treatments that combined thinning and prescribed burning. Prescribed fire emissions correlated with stocking levels, with the burn-only having the highest C release and the 1865-burn treatment having substantially lower emissions (Figure 2). The fire emissions pattern is the same when wildfire emissions are included, with all prescribed-burn treatments having higher totals than their unburned, paired treatment. Harvested trees accounted for 65 t C ha–1 and 47.8 t C ha–1 removed from the site for the understory and restoration thinning treatments, respectively. This C, however, is not emitted directly to the atmosphere. For a merchantable timber sale such as Teakettle, about 40% may become rapidly decomposed milling “waste” such as sawdust, and 60% can become wood products with a half-life of 1–100 years (Skog and Nicholson 2000).
Thinning trees from small size classes had little impact on tree-based C storage, but did raise the average height from the ground to the base of the live crown, a key factor in reducing fire intensity (Agee and Skinner 2005).
Long-term fuels reduction was greatest for restoration and prescribed-burn treatments, because they reduced small-diameter tree densities and shifted composition toward more fire-resistant pines. The most wildfire-resistant treatments, as measured by those with the highest number of large trees (> 75 cm dbh), were the 1865, 1865-and-burn, restoration-and-burn, and burn-only treatments, whereas the control and understory-thin treatments had the fewest (Figure 3).
Discussion
In flammable forests, sequestering C is more complex than maximizing stocking levels and mean annual growth increments. There are trade-offs in emission and storage rates, depending on treatment application and wildfire timing as stands develop. However, the consistently high storage and low emissions of the 1865 reconstruction suggest that a low-density forest, dominated by large, fire-resistant pines, may be a desired stand structure for stabilizing tree-based C stocks in wildfire-prone forests. …
Recent research suggests that immediate wildfire emissions may only be a portion of actual C losses, if the fire leaves few surviving trees (Kashian et al. 2006). Auclair and Carter (1993) calculated that high-intensity, post-wildfire C release was approximately three times the direct release of CO2 during the fire event. In ponderosa pine, direct flux measurements found higher CO2 emissions from a high-intensity burn than those from an unburned site, even 10 years after fire (Dore et al. 2008). Future research may more effectively incorporate these C losses associated with high-intensity fire into models, but, in this paper, we compare only direct C emissions occurring during the fire. …
The low density of the 1865 reconstruction consolidates increment growth in large, fire-resistant trees, while maintaining fewer small trees (Figure 3). The restoration-and-burn treatment (Figure 3f) is the best option for approximating the 1865 forest structure and species composition, conditions that should be fire resistant. We caution, however, that our modeling focus is on stands exposed to a simulated, uniform wildfire event. Wildfire effects on forest conditions and C emissions will vary across a burn landscape in response to local fuel conditions and the interaction of fire behavior and weather. We have not attempted to model this more complex fire dynamic and instead have focused on the scale at which managers often manipulate forest structure and use different fuel treatments. …
Forest C sequestration has been proposed as a way to help offset other anthropogenic CO2 emissions (Woodbury et al. 2007). In forests that historically burned with high frequency and low severity, adding to the C baseline by increasing stocking levels may exacerbate the modern shift toward high-severity fire produced by fire suppression and climate change. Current C accounting practices can be at odds with efforts to reduce fire intensity in many western US forest types. Although the concept of restoring forests in the western US to some pre-settlement target may not be feasible as the climate changes, reducing fire severity and increasing and stabilizing tree-based C storage may be achieved with fuel treatments that promote low-density, large pine-dominated stand structures.
