Post‐Fire Comparisons of Forest Floor and Soil Carbon, Nitrogen, and Mercury Pools with Fire Severity Indices

Abstract

Forest fires are important contributors of C, N, and Hg to the atmosphere. In the fall of 2011, a large wildfire occurred in northern Minnesota and we were able to quickly access the area to sample the forest floor and mineral soil for C, N, and Hg pools. When compared with unburned reference soils, the mean loss of C resulting from fire in the forest floor and the upper 20 cm of mineral soil was 19.3 Mg ha−1, for N the mean loss was 0.17 Mg ha−1, and for Hg the mean loss was 9.3 g ha−1. To assess the influence of fire severity on the forest floor and mineral soils, we used an established method that included a soil burn severity index and a tree burn severity index with a gradient of severity classes. It was apparent that the unburned reference class had greater forest floor C, N, and Hg pools and higher C/N ratios than the burned classes. The C/N ratios of the 0‐ to 10‐ and 10‐ to 20‐cm mineral soils in the unburned reference class were also greater than in the burned classes, indicating that a small amount of C was lost and/or N was gained, potentially through leaching unburned forest floor material. However, with a couple of exceptions, the severity classes were unable to differentiate the forest floor and mineral soil impacts among soil burn and tree burn severity indices. Developing burn severity indices that are reflective of soil elemental impacts is an important first step in scaling ecosystem impacts both within and across fire events.

Soil organic matter composition and quality across fire severity gradients in coniferous and deciduous forests of the southern boreal region

“…We found that fire caused forest floor thickness, C concentration, and C stocks to decline, whereas there was no effect of fire on mineral soil C or N concentrations or stocks, across coniferous and deciduous forest cover types. These results agree with estimates from an earlier, broader evaluation at this study site that included a greater range of forest cover types [Kolka et al, 2014]. However, we observed a decrease in C:N ratios in forest floor and mineral soil layers in all severity levels even when there were no significant differences in C or N stocks.…”

“…Upland soils are entisols classified as well‐drained, shallow (20–50 cm) gravelly coarse sandy loam over bedrock (Quetico (Lithic Udorthents) and Insula (Lithic Dystrudepts) series) or moderately deep (50–100 cm) gravelly sandy loam over bedrock (Conic (Typic Dystrudepts) and Wahlsten (Oxyaquic Dystrudepts) series) (J. Barott, Soil Scientist, USFS Superior National Forest, personal communication). A more detailed description of the study site and the Pagami Creek wildfire has been presented previously [ Kolka et al ., ].…”

“…Scientist, USFS Superior National Forest, personal communication). A more detailed description of the study site and the Pagami Creek wildfire has been presented previously [Kolka et al, 2014].…”

“…As part of the larger investigation [Kolka et al, 2014], we used knowledge of prefire vegetation [Wolter and Townsend, 2011;Wolter et al, 2009] and remotely sensed estimates of fire severity (relative differenced normalized burn ratio) [Miller et al, 2009] to establish a series of transects and plots to capture gradients in expected fire severity. However, fire severity was assessed in the field (see details below), and these fieldbased severity estimates were used to designate the classifications we report in this study.…”

Quantifying Early-Seral Forest Composition with Remote Sensing

“…This mapping approach can further compliment more detailed field studies of vegetation dynamics by providing vegetative responses across the full range of environmental conditions, characteristic of large landscapes . Future research should address vegetation patch characteristics and scale, as well as species-specific associations and communities in terms of competition with forest regeneration (Frey et al, 2003), and how such associations might correspond with certain pre-fire cover types and structure as well as post-fire soil characteristics and fire severities (Kolka et al, 2014).…”