Effects of Multiple Interacting Disturbances and Salvage Logging on Forest Carbon Stocks

Forests contain approximately 80% of aboveground terrestrial carbon. Therefore, minor alterations to carbon stocks or cycling in forests ecosystems may exacerbate global warming by increasing atmospheric carbon dioxide levels. Yet, current climate change is expected to increase the severity and frequency of stand-replacing disturbances such as wildfire and windthrow, which will ultimately decrease ecosystem carbon stocks over large areas for several decades. Previous studies have assessed the effects of individual disturbance on forest carbon storage. However, the consequences of multiple, interacting disturbances are relatively unknown. Using field data and statistical analyses, Bradford et al. (2011) quantified the changes in carbon stocks in ecosystem carbon pools (live biomass, snags, down woody debris, forest floor, and total ecosystem) that resulted from blowdown from windstorms, post-disturbance salvage logging, and wildfire in a forest in Northern Minnesota. Each disturbance was analyzed individually and in combination with one another. The authors found that total live carbon and carbon in live trees were highest in the control plot (no disturbance) and lower in all other treatments. Carbon in understory biomass was highest in the blowdown followed by salvage logging followed by fire plot (BSF). Carbon in snags and dead woody material was highest in the fire treatment while carbon in down woody debris and the florest floor was highest in the blowdown treatment. Total ecosystem carbon was highest in the control treatment, intermediate in the blowdown and fire treatment, and lowest in the blowdown and fire (BF) and BSF plots. Both blowdown and fire disturbances resulted in roughly equal decreases in live carbon and total ecosystem carbon. Fire further decreased carbon in the forest floor and down woody debris after blowdown, resulting in additional total ecosystem carbon losses. Salvage logging and fire after blowdown demonstrated similar results. Overall, these results indicate that increasing disturbance frequencies may challenge land management efforts to sustain and enhance ecosystem carbon stocks.—Megan Smith

Bradford, J.B., Fraver, S., Milo, A.M., D’Amato, A.W., Palik, B. Shinneman, D.J., 2012. Effects of Multiple Interacting Disturbances and Salvage Logging on Forest Carbon Stocks. Forest Ecology and Management (267), pgs. 209 – 214.

The study was conducted in Northeastern Minnesota along the Southern edge of the North American boreal forest ecotone, and within the Superior National Forest. On July 4, 1999, a large derecho (a widespread, straight-line windstorm) affected over 200,000 hectares of the Superior National Forest. After the storm, the US Forest Service began salvage logging operations to reduce fuel loads and mitigate wildfire. Salvage logging occurred between the fall of 1999 and the fall of 2002. Despite these efforts, a section of the region was burned by the Ham Lake Wildfire in 2007. Patchy, overlapping disturbance patterns resulted in five “treatments:” undisturbed control, blowdown only (B), fire only (F), blowdown followed by fire (BF), and blowdown followed by salvage logging followed by fire (BSF). Unburned salvaged areas were unidentifiable. The authors examined mature jack pine (Pinus banksiana) communities. A map displaying the location of the 1999 blowdown, the 2007 wildfire, and the study site was constructed.

Bradford et al. established six study sites in each study plot, creating 30 sites total. Eight of these sites contained previous data from an earlier investigation and were included in the author’s study. The remaining 22 sites were randomly selected. Although the authors lacked pre-disturbance data for most of their sites, they conducted comparisons of treatment stand structures and reconstructed structures for the other treatments (based on deadwood pools) with the plots with pre-existing data. As a result, the authors confirmed that all the sites were comparable regarding pre-disturbance stand structure and successional stage. Randomly selected sites were then ground-truthed for adherence to the expected forest type and disturbance combination. Within each site, 6–10 circular plots (200-m2) were established along a 40 x 40 m grid that originated from a randomly chosen starting point, creating 239 plots total.

Within each circular plot, standing live and dead trees (diameters greater than 10 cm at breast height), and saplings (stems sizes greater than 2.5 cm and diameters less than 10 cm at breast height) were recorded by species and diameter. Stems of shrubs and tree seedlings (stems smaller than sapling class) were tallied by species within a 10-m2 circular plot centered within each 200-m2 plot. Additional seedling data were collected in in 10-m2 plots located equidistant between each 200-m2 plot, resulting in 14 – 20 seedling plots at each site. Downed woody debris (DWD) on each plot was inventoried. Diameters of all DWD with measurements greater than 7.6 cm were recorded along a 32-m transect that passed through the middle of each 200-m2. The authors also collected samples of herbaceous vascular plant material, forest floor material, and soils at a set location within each 200-m2 plot.

The biomass of living and intact standing dead trees was calculated for all woody stems greater than 2.5 cm in diameter and breast height. The authors used species-specific allometric equations that were regionally derived. The biomass of broken standing dead trees was estimated using taper functions to determine large and small end diameters. Then, a conic-paraboloid formula was applied to determine the volume of the intact portion of the tree. Volume was converted to biomass using species-specific density values for decay class I taken from another author’s study. For unidentifiable standing dead trees, the authors used the average decay class II densities from all species present. Shrub and tree seedling biomass was calculated using species-specific allometric equations. Biomass calculations for these components included stems, roots, branches, as well as foliage, if it was alive. Dead woody debris biomass was calculated using Van Wagner’s formula. This formula was applied to the planar intercept data, adjusting for species and decay class-specific densities. The volume of coarse woody debris in decay class IV or V was adjusted for collapse using collapse ratios of 0.82 and 0.42. Carbon content was calculated from total biomass using species-specific values. Additionally, carbon was assumed to compose 50% of biomass for shrub species and unidentified down woody debris.

Separate mixed-model analyses of variance (ANOVA) were used to assess differences in carbon pool sizes among treatments. Disturbance combination was treated as a fixed effect and site was treated as a random effect. Analyses of variance was also calculated for carbon stored in live trees, understory (herbaceous plants, shrubs, and seedlings), live biomass (sum of live trees and understory), down woody debris (DWD), snags, dead woody material (snags plus DWD), forest floor, soil, and total ecosystem carbon (sum of live biomass, deadwood, forest floor, and soil). When the authors detected significant disturbance effects, post hoc Tukey’s honest significant difference tests were used for pairwise comparisons between disturbance types. All data were checked for normality and transformed before statistical analyses were undertaken.

Brandon et al. found that carbon in live biomass was stored primarily in live trees. Total live carbon and carbon in live trees were highest in the control plot (104 MgC ha-1) and lower in all other treatments (from 1 MgC ha-1 in the BSF plot to 21 MgC ha-1 in the blowdown treatment). Carbon in understory biomass was higher in the BSF treatment (1.3 MgC ha-1) than the control, blowdown, or fire treatments (0.2 to 0.6 MgC ha-1). Carbon in snags was much higher in the fire treatment than other treatments (53 MgC ha-1), and the BF treatment (19 MgC ha-1) was higher than the BSF treatment (1.7 MgC ha-1).

Carbon in DWD was highest in the blowdown stands (35 MgC ha-1), intermediate in the BF treatment (21 MgC ha-1), and lowest in all other treatments (9–10 MgC ha-1). Carbon in all dead woody material (snags and DWD) was highest in the fire treatment (62 MgC ha-1) and lowest in the BSF treatment (11 MgC ha-1). Carbon stored in the forest floor was highest in the blowdown treatment (35 MgC ha-1) and lowest in the BSF treatment (9 MgC ha-1). Furthermore, soil carbon means were not significantly different among treatments. Total ecosystem carbon was highest in the control treatment (177 MgC ha-1), intermediate in the blowdown and fire treatments (122 and 106 MgC ha-1), and lowest in the BF and BSF treatments (66 and 40 MgC ha-1).

Differences in carbon pools suggest that individual natural disturbances (blowdown and fire) resulted in similar decreases in live carbon (83 and 91 MgC ha-1) and total ecosystem carbon (55 and 71 MgC ha-1). Blowdown increased DWD carbon by 25 MgC ha-1 and forest floor carbon by16 MgC ha-1 and had no significant effect on sang carbon. However, fire increased snag carbon by 40 MgC ha-1. Compared to blowdown alone, fire following blowdown decreased both DWD and forest floor carbon by another 14 and 22 MgC ha-1, and resulted in additional total ecosystem carbon losses of 56 MgC ha-1. Similarly, salvage logging and fire after blowdown decreased carbon in both DWD and forest floor by another 26 and 39 MgC ha-1, and caused additional ecosystem carbon loss of 83 MgC ha-1 when compared to blowdown alone. A figure displaying carbon stocks for individual carbon pools and total ecosystem carbon for forest stands in each treatment was constructed, as was a figure demonstrating the consequences of windthrow, salvage logging, and wildfire on ecosystem carbon stocks.

These findings indicate that all combinations of disturbances resulted in lower total ecosystem carbon than the undisturbed control. Additionally, the results demonstrate that individual disturbances (wind and fire) resulted in similar shifts of carbon from live biomass to detrital pools, as well as similar losses of total ecosystem carbon. However, blowdown shifted live tree carbon into downed woody debris and forest floor pools while fire shifted this carbon into snags, which have slower decomposition rates than downed wood or forest floor. Total ecosystem carbon loss was attributed to the amount of carbon immediately lost from trees.

Overall, the authors’ results demonstrate that secondary major disturbances can cause substantial additional decreases in ecosystem carbon, though the magnitude may vary depending on the time between successive disturbances and disturbance types. Furthermore, the original blowdown disturbance increased carbon stored in DWD and forest floor, providing larger surface fuel loads that increased burn severity and carbon loss. Although forest floor and DWD carbon pools were elevated after the blowdown, the fire released all the carbon added to the litter in the forest floor and more than half of the carbon added to DWD. Salvage logging also modestly enhanced carbon losses. Post-blowdown salvage logging prior to fire decreased carbon in DWD to control levels, removing all of the carbon added to DWD in the blowdown. It also resulted in less carbon stored in snags. Yet, salvage logging did not significantly affect total ecosystem carbon despite its effects on individual carbon pools.

In conclusion, forest managers must address the issue that increasing disturbance frequencies may counteract efforts to sustain and enhance ecosystem carbon stocks, thereby exacerbating climate change by releasing more carbon into the atmosphere.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s