97th ESA Annual Meeting (August 5 -- 10, 2012)

PS 104-178 - Patch characteristics of forest tornado damage in rugged terrain

Friday, August 10, 2012
Exhibit Hall, Oregon Convention Center
Jeffery B. Cannon1, Jeffrey Hepinstall-Cymerman2, Luke J. Snyder1 and Chris J. Peterson3, (1)Department of Plant Biology, University of Georgia, Athens, GA, (2)Warnell School of Forestry & Natural Resources, University of Georgia, Athens, GA, (3)Dept. of Plant Biology, University of Georgia, Athens, GA
Background/Question/Methods

Large gaps created by major forest disturbances influence numerous community and ecosystem characteristics and processes. Although wind is a primary disturbance agent in many forests, only a few studies have examined landscape-scale patterns of wind damage in forests, and to date, all those have been based on damage from hurricanes, downbursts, or unusally low jet stream in mountain areas. Gap characteristics and spatial patterns of severity have not been studied at the landscape scale for tornado‐damaged forests. While it is expected that topography influences tornado damage patterns, this expectation has not been examined quantitatively. We hypothesized that topography caused consistent patterns of variation in blowdown severity following a tornado in the Southern Appalachians. On April 27, 2011 a tornado damaged mixed oak-pine forests of northern Georgia in the Chattahoochee National Forest. The objective of this study is to use aerial photographs of damaged areas to estimate spatial distribution of severity and to characterize blowdown patches created by tornado damage.

Aerial photos (20 cm resolution) were taken shortly after the tornado. Images encompassed the entire width of the tornado damage and extended several hundred meters into intact forest on either side. We measured severity (relative basal area down) in thirty-eight 20x20 m plots. These plots were used as training sites in Supervised Maximum Likelihood Classification which estimated damage severity across the entirety of the aerial photograph. We used multiple regression to assess how tornado severity is influenced by aspect, slope, and elevation using data from a digital elevation model of the study area (30 m resolution). Lastly, we grouped contiguous patches of damage to assess patch characteristics such as number and size-distribution.

Results/Conclusions

We found that damage severity was greatest in areas proximal to the tornado path, on exposed aspects, on steeper slopes, and on lower absolute elevations indicating that severity is strongly influenced by topographic variation. We found that tornado damage created heterogeneous patch sizes ranging in size from 0.04 ha to greater than 60 ha. Although small patches were much more common than larger patches, the latter accounted for a large majority of the damaged area. Of all areas with damage (>20% basal area down), low severity damage (21-40% of basal area down) covered 57.3% of the area. Moderate severity damage (41-60% severity) covered 31.2% of the area. High severity damage (>60% severity) covered 11.5% of the area. Thus low severity damage was most widespread.