PARENT SESSION
Contributed Oral Session 107: Soil Biota and Communities
Wednesday, August 10, 1:30 PM - 5:00 PM, Meeting Room 524 B, Level 5, Palais des congrès de Montréal

Spatially-explicit modeling of soil respiration rate in a high-elevation, subalpine forest.

Scott-Denton, Laura^*,1, Monson, Russell¹, ¹ University of Colorado-Boulder, Boulder, CO, USA

ABSTRACT- Ecosystem level estimates of soil respiration are difficult to generate due to large spatial heterogeneity in soil respiration rates, even at small scales. Many researchers have used plants to predict spatial patterns in soil processes, but this approach is complicated in a forest environment where any point on the soil can be influenced by multiple trees. This study investigates an index of tree influence on soil (V) than can be used as a proxy for soil respiration rate and can easily be scaled up to a canopy estimate of soil respiration in a subalpine coniferous forest at the Niwot Ridge Ameriflux Site in Colorado, USA. First we demonstrate that V predicts soil respiration rate by using a sample of fifty point-level soil respiration rate measurements and manually mapping surrounding trees to calculate V. Then we use Global Positioning Satellite (GPS) technology to map every tree within sixteen 10 x 10 m plots forming a transect of plots from the east to the west of the eddy covariance tower at the site. Using Geographic Information System (GIS) software, each mapped plot is then superimposed by a 10 x 10 cm grid, and a value for V is calculated for each grid cell. The maps of V are then used to estimate soil respiration for the site. This modeled estimate for soil respiration generated by this spatially explicit method of scaling up agrees well with ecosystem respiration fluxes predicted by nighttime eddy covariance at the site (3.23 mol m-2s-1 compared to 4.33 mol m-2s-1). Respiration rate, tree density, and species composition all change over the east-west gradient, as the Niwot Ridge Ameriflux tower is placed on the ecotone between higher-elevation spruce-fir forest to the west and lower-elevation lodgepole pine dominated forest to the east. Multiple regression analysis reveals that it is tree density, not species composition that is driving the gradient in soil respiration. At the point-level collar measurements of respiration, local tree density could not predict soil respiration rate, suggesting that this relationship may be only observable at larger (plot-level or greater) spatial scales. Tree density may be a useful proxy for estimating soil respiration at regional scales.

Key words: soil respiration, stem maps, GIS, conifer forest